Full text of "Kluwer - Compact Disc Interactive Designer's Overview - Second Edition"

See other formats

_ COMPACT DISC-INTERACTIVE
_ ADESIGNER’S OVERVIEW |

Edited by Philips International - J.M. Preston Kluwer

ee ee a — " 2 eee <a

Compact Disc-Interactive:
A Designer’s Overview

Compact Disc-Interactive
A Designer’s Overview

Edited by J.M. Preston of Philips International in Eindhoven

Foreword by G.A.J. Bastiaens, Director Consumer Elec-
tronics, Interactive Media Systems, Philips International BV
and G. Stulberg, Chairman & Chief Executive Officer,
American Interactive Media Inc.

Kluwer Technical Books
Deventer - Antwerpen

Distribution:

Kluwer Technical Books
P.O. Box 23

7400 GA Deventer

The Netherlands

ISBN 90 201 21219

First edition, December 1987
Second edition, March 1988
Second printing, November 1988

All rights reserved. No part of this book shall be reproduced, stored in a
retrieval system, or transmitted by any means, electronic, mechanical,
photocopying, recording, or otherwise, without written permission from
the publisher. No patent liability is assumed with respect to the use of
the information contained herein. While every precaution has been taken
in the preparation of this book, the publisher and the author assume no
responsibility for errors or omissions. Neither is any liability assumed
for damages resulting from the use of the information contained herein.

FOREWORD

Compact Disc is without doubt one of the most successful products in
the history of Consumer Electronics. Of the many reasons for this success
the following three are the most important:

* The specific properties of the CD system e.g. compact in size, easy to
handle digital encoding, perfect sound, fast access, massive storage
capacity, as well as its ability to be used in many different environ-
mental conditions.

¢ A comprehensive and attractive disc software catalogue.

The same is true for the newly introduced CD-Video where high quality
video is combined with CD sound. A third successful product in the
optical family is the CD-ROM, mainly used as a computer peripheral for
fast retrieval of text and data. The combination of high quality audio with
video, graphics, text and data, all interleaved together and offering fast
information retrieval under user control, has created a new and fas-
cinating product, called CD-Jnteractive or CD-I.

CD-I offers a degree of flexibility in the presentation of information
never before possible in consumer products, bringing a new dimension
to the ways in which information will be published, distributed and used.
It combines the power of a computer system for fast access to in-
formation, with the excitement of attractive multimedia software for
entertainment, education and information in such a way that the con-
sumer is unaware that there is a computer involved. He or she just enjoys
the application on the disc.

It is self evident that music is an interesting source of software for CD-I,
combining music with the display of high quality pictures, or even
motion pictures, together with text and data. We believe that this will
create many enjoyable applications. Adding sophisticated new video and
audio enhancements to adventure games or games of skill will create an
interesting new range of entertainment software.

The need to improve the quality of education both in the home and school
environment will be one of the most universal priorities in society during
the next decade and there CD-I, as a truly multi-media system, can offer
the type of breakthrough required. CD-I will help the user keep up to date
with the latest developments in his or her area of activity, setting the pace
according to the user’s level of knowledge. Reviewing or re-checking of

the source material by virtue of CD-I’s fast access capability becomes
much easier and more exciting than browsing through a book. Even
existing publications, suchas, for example, an encyclopedia, will become
a completely new product realizing the full flexibility and potential
offered by CD-I.

As we consider CD-I principally as a software driven mass consumer
multi-media system, its success depends entirely on the attractiveness
and variety of the software catalogue. This software catalogue will
contain video enhanced music, games, sports, reference books, self help,
children programs, all kind of educational material as well as tourist and
travel information.

Alongside the very interesting consumer and educational market CD-I
will create attractive possibilities for the institutional and professional
area, typically in point of sales, point of information and training
applications.

One of the critical factors of success for CD-I is the availibility of
qualified authoring systems. For this, several possible authoring en-
vironments are available, such as Microware, Microboards, Sun, and
Optimage. Alongside these systems there are prototype CD-I players
which will be available to form part of low cost authoring equipment.

Sony and Philips have now completed the CD-I technical specifications,
known as the "Green Book Standard". The hardware and software
industries are now actively involved with the preparation for the in-
troduction of CD-I.

Many software companies are already preparing software for CD-I. Our
own American Interactive Media Company (AIM) is preparing the
software catalogue for the US market, together with many other software
providers, while similar developments are also in progress in Europe
(EIM) and Japan (JIM). We are convinced that other companies will
become part of this fascinating development.

We firmly believe that this book will give an excellent introduction to
the CD-I software creation environment, because "the software is the key
to success".

G.A.J. Bastiaens G. Stulberg
Director Consumer Electronics Chairman & Chief Executive Officer
Interactive Media Systems American Interactive Media Inc.

PREFACE TO SECOND EDITION

This book is an introduction to CD-I. It has been written to provide an
overview of the current state of development for those who want to know
more about the topic - perhaps with an ambition to become CD-I
publishers or to become CD-I designers.

It is not a designer’s guide. That is, it cannot be used as a vademecum
while designing a CD-I disc. At the time of writing there is an insufficient
body of experience to enable such a guide to be written - only two CD-I
discs have been displayed to Licensees, one by Sony in June 1987 and
one by Philips in December 1987. The book does, however, represent
the cumulative experience of most of those already involved in CD-I.

It is likely that the book will go through many revisions as the body of
experience grows and many of our expectations are either confirmed or
denied. The long term success of CD-I is going to be determined by the
foresight of the publisher and the ingenuity of the designer; so the more
experience can be shared at this stage, the greater the effort that can go
into original design work. The reader is therefore invited to write to the
editor (c/o the publisher) with comments and criticisms so that subse-
quent editions may be both more informative and more comprehensive.

This second edition has been edited by Dick Fletcher of New Media
Projects in London. The substantial revisions to the first edition have
been made possible in particular by the guidance and detailed comments
of:

Richard Bruno of Optimage in Chicago

Peter Cook of Grolier Electronic Publishing in New York
Peter Essink of Japan New Media Systems in Tokyo
Nicholas Lewis of New Media Projects in London

Graham Sharpless of Home Interactive Systems in Eindhoven

J.M. Preston
Philips International
Eindhoven

1 March 1988

Table of Contents vii

TABLE OF CONTENTS

HOW TO USE THIS BOOK /

CHAPTER 1: INTRODUCTION 3

CD-I TRUE MULTI-MEDIA TECHNOLOGY 3
WHAT IS MULTI-MEDIA? 3

Why is Multi-media important? 3

Multi-media CD-ROM 4

CD-ROM XA 4

LaserVision 4

CD-I and Multi-media 5
BEYOND MULTI-MEDIA: HYPERMEDIA 5
INTERACTIVITY - THE ’I’ INCD-I 6

What is Interactivity? 6

How to use interactivity? 6

Interactivity : The Educator 7

Do Consumers Want Interactive Programs? 7

CHAPTER 2: THE BACKGROUND OF CD-I 9

THOSE CHILLY SATURDAYS 9
ELECTRONIC PUBLISHING 1/1

OPTICAL DISCS /2

COMPACT DISC (CD) /5

COMPACT DISC - DIGITAL AUDIO (CD-DA) /5
ANALOG AND DIGITAL 16
INTERNATIONAL STANDARDS 16
COMPACT DISC - INTERACTIVE (CD-I) 17
THE CD-I PLAYER 17

CD-I AUDIO /8

CD-I VIDEO /8

THE DESIGN PROCESS 19

CHAPTER 3 : WHAT CD-ICAN DO 2]

USING AN INTERACTIVE TELEVISON SET 2/
AN EXAMPLE : CD-IGOLF 22
AUDIO 23

Quality Levels and Capacity 23

viii CD-l: A Designer's Overview

Soundmaps 25
Specialist Use 26
Audio Control 26
VIDEO 26
National TV Broadcast Standards 26
Resolution 26
Safety Area 27
Compression 28
Coding Techniques 29
Image Planes 31
Motion 32
VISUAL EFFECTS 34
Single plane operations 34
Cuts 34
Sub-screens 35
Scrolling 35
Mosaic Effects 36
Fade 38
Two Plane Effects 38
Transparency 38
Chroma Key 38
Mattes 39
Transparent Pixels 39
Dissolves 40
Wipes 40
REAL-TIME INTERACTIVITY 41
Real-Time Data 41
Synchronization 4/
USER INTERFACES 42
Physical Interface 42
Interacting with the user 43
CONCLUSION 43

CHAPTER 4: THE DESIGN BRIEF 45

DEVELOPING THE DESIGN BRIEF 45
DEVELOPING THE IDEA MAP 52
TREATMENT 54
THE DESIGN TEAM 54

A Typical Design Team 55
BUDGET AND SCHEDULE 56

The Production Process 57
CONCLUSION 59

Table of Contents ix

CHAPTER 5 : DESIGNING FOR PRODUCTION 1 6/

DEVELOPING THE STORYBOARD 6/
THE CD-I AUTHORING ENVIRONMENT 61]
The Designer’s Station 63
The Production Facility 63
Disc Building 64
Standard Data Formats 64
Scripting Subsystem 64
Audio Subsystem 65
Video Subsystem 65
The Presentation Editor 65
The Database Subsystem 65
Testing and Simulator Subsystem 66
BASIC PRINCIPLES 66
Tracks and Sectors 67
Disc Capacity 68
Data Transfer Channels 69
Data Transfer Channels: Audio Playback 70
Data Transfer Channels: Video Playback 7/
System RAM 72
Microprocessor 73
THE MECHANICS OF CD-I DESIGN 74
Interface Devices 74
Screen Design 74
Screen Effects 75
Controlling the Dataflow 77
Interleaving 77
Seek Time 79
Synchronization 80
Synchronizing to Audio 80
Synchronizing to Video 81
Timed Cues 8/
Interactive Design 8/
CONCLUSION 82

CHAPTER 6 : TYPICAL CD-I APPLICATIONS 83

THE GROLIER MULTI-MEDIA ENCYCLOPEDIA 85
Designing the Interactivity 85
Mode of use 85
Data Management 86
Screen Interface 86
Menus 86
Screen Effects 87

x CD-I: A Designer's Overview

Graphic Control Panel 87
Interactive Branches 87
Domain Seeks 88
HOT SHOT SPORTS 88
Photographic Database 89
The Animated Figure 92
Action Areas 93
Audio 94
COUNTRY HOUSE MURDERS 94
Surrogate Travel 94
Handling Disc Space 95
Icons 95
Audio 96
POP SHOWCASE 96
Partial Screen Updates 97
Disguising Disc Seeks 97
Singalong 98
Screen Proportion 98
INTERACTIVE UNDER FIVES 99
Shape Recognition 99
FRENCH PHRASEBOOK 100
Chroma-key Facility 102
Real-time Dramatizations 102
CONCLUSION 107

CHAPTER 7 : HOW CD-I WORKS 109

DISC STRUCTURE 109
Disc Organization 109
Disc Label 110
Path Table and Directories J/1
Files 112
CD-ISECTORS //2
CDI Sector Format 112
CD-I Audio Sectors 113
CD-I Video Sectors 114
DYUV Images 115
CLUT Images 1/5
RGB 5:5:5 Images 116
Run-Length Images 116
Program Related Data sectors 117
CD-I DECODER 117
CD-Drive/Player 118
Audio Processor 120
The Audio Processing Unit 121

Table of Contents xi

Video Processor 12]
RGB Levels 125
Display Control Program (DCP) 126
The Hardware Cursor 128
Border Color 128
DMA Controller 129
Random Access Memory (RAM) 129
Non-volatile RAM (NV-RAM) 1/30
Clock/calendar 130
Pointing Device 130
Transfer Paths 130
Application Software 130
Audio Pathways 132
Playback from System Memory 1/32
Video Pathways 132
Image Store (RAM) 1/32
CD-RTOS 133
Organization 133
THE KERNEL 1/35
Configuration Status Descriptor 135
Start-up Procedure 135
File Protection 1/37
FILE MANAGERS 137
UCM Video 137
UCM Audio 139
UCM User Interface 139
Compact Disc File Manager (CDFM) 139
NV-RAM File Manager (NRF) 140
DRIVERS 140
SYNCHRONIZATION AND CONTROL 1/40
Play Control Structure /4/
Real-time Control Area 14]
InVision 142

APPENDICES
A: TECHNICAL SPECIFICATION SUMMARY 1/43

B: AGLOSSARY OF TERMS 1/45
C: CD-RTOS AND INVISION 20]

INDEX 237

HOW TO USE THIS BOOK

This guide has been designed to accommodate a variety of readers’ needs
and interests. It is anticipated that some people will need to study in
depth, while others will want only an introduction. Even those who will
eventually study the entire guide may read selectively at first, to gain a
solid background before coming to grips with the finer details.

The book grows more technical as it progresses. The seven chapters
break down this way:

Chapter 1 explains the concepts of multi-media and interactivity.

Chapter 2 offers a general background to electronic publishing and
optical disc technology. The CD-I player and disc are described, together
with the principal audio-visual features of the system.

Chapter 3 looks at the audio-visual side of the technology in more detail,
and explains what CD-I can do, and the design implications of its many
features.

Chapter 4 looks critically at the kind of questions the potential designer
must ask before undertaking a CD-I project and designing the brief.

Chapter 5 discusses the design process in some detail, and describes
essential stages and production tasks.

Chapter 6 presents a range of hypothetical projects which might be
among the first CD-I discs on the market: a multi-media encyclopedia,
a pop music program, a language program, and several different game
styles. Some of the various features and design considerations unique to
CD-I are illustrated by specific examples within each application.

Chapter 7 discusses the technology in depth, from a computing
perspective, and describes elements of the player and the technical
composition of the disc itself.

While the book flows logically from Chapter 1 to Chapter 7, readers with
different needs and interests may want to approach the guide itself
interactively’.

Chapters 1 and 2 provide a general introduction and are appropriate for
senior management and strategists.

How to use this book

2 CD-Il: A Designer's Overview

Itis likely that the potential CD-I designer or producer will read the whole
book, but may wish to begin with the first two chapters for a conceptual
introduction, and skip to Chapter 6 to look at how typical applications
might work. Many may then prefer to think about the design process
described in Chapters 4 and 5 before tackling the technical material.

What to look for
© Chapter 1: multimedia and interactivity

Chapter 2: optical publishing panorama

Chapter 3: the media palette

Chapter 4: some pre-design questions

Chapter 5: design considerations

Chapter 6: some CD-I examples

Chapter 7: inside the technical system

Appendices:
© Glossary
° CD-RTOS and InVision
° Index

Because this guide covers so much inter-related material - such as the
description of a feature from both a design and a technical point of view
- it contains aids to help readers cross-reference new ideas and concepts
as they appear in different contexts. These include:

© Side comments to help identify cross references.
¢ A Glossary explaining key words and concepts at the end of the book.

CHAPTER 1: INTRODUCTION

This chapter introduces the two basic concepts behind Compact
Disc-Interactive (CD-1), the most recent development in the tradition of
optical disc recording - multi-media and interactivity.

CD-I: TRUE MULTI-MEDIA TECHNOLOGY

Compact Disc-Interactive (CD-I) will be the first publishing medium to
bring the world of multi-media to a broad general audience. It is essential
for the CD-I publisher, designer and producer to understand this concept,
and why CD-I is so much better suited to multi-media applications than
other technology.

WHAT IS MULTI-MEDIA?

The term multi-media originated with the audio-visual industry, to
describe a computer-controlled, multiple-projector slide show with a
sound track. In computer terms, multi-media is viewed as a blending of
media types: text, audio, visual, and computer data in one convenient
delivery medium. Although CD-I is not the only combination of
hardware and software capable of delivering multi-media information,
it is the first to do so in a highly standardized form - and, for broad
acceptance of a technological concept such as multi-media, system
standards are a prerequisite. CD-I has been defined as a system standard,
in contrast to CD-ROM which functions simply as a peripheral to a
system. The CD-I specification defines how the information is stored on
the CD-I optical disc, exemplifies how it is encoded in the recording
studios, and defines how it is decoded in the player. A peripheral such
as CD-ROM defines only how the information is stored on the disc,
making international agreement on recorder and player standards
impossible.

Why is Multi-media Important?

Information sources such as books, periodicals, film, television, radio,
video, LPs, cassettes, and computer software, have evolved along various
tracks and, in our minds, are viewed as separate and distinct media. But
information need not be defined by the medium in which it is presented.
The development of CD-I enable us for the first time to mix information
from a variety of sources, using the medium most appropriate to the
message - a short sequence of images called a video clip, computer
animation, or a screen full of text, all supported by any combination of
speech, music and sound effect as needed. CD-I combines the best

Chapter 1: Introduction 3

4 CD-l: A Designer's Overview

creative concepts from book design, film and video production, sound
recording, and software design. This represents a new range of
challenges for the designer, who must select the most effective
combinations from the rich media palette of CD-I.

Multi-media CD-ROM

CD-I is an optical disc technology, the logical extension of both the
Compact Disc and Interactive LaserVision. In 1985, CD-ROM
(Compact Disc-Read Only Memory) was introduced as a mass storage
peripheral for personal computers (PCs). Though developed primarily
for text, it can store digital data of any kind, including sound and
graphics. However, the multi-media potential of CD-ROM is defined by
processing power, audio output, and the display capabilities of the
computer that is controlling the CD-ROM drive. While most PCs can
generate adequate audio and graphics output for computer applications,
they were not designed to produce high quality audio or to display natural
video images. CD-ROM is nevertheless well suited to dedicated
solutions, where internationally agreed standards do not play a
significant role.

CD-ROM XA

N.V. Philips, Sony Corporation and Microsoft recently together
announced the joint development of an extended, multimedia format for
CD-ROM, known as CD-ROM XA. The CD-ROM XA development is
based on the following approach:

® Interleaved audio in ADPCM format, as defined in the CD-I format.

® Interleaved text/graphics, based on specified screen format for perso-
nal computer displays.

CD-ROM XA will provide a variety of functions which are similar to
those found in the CD-I format, but which will not be dependent on a
specific software operating system or microprocessor. This new format
will allow publishers to create discs which will be able to be played on
any suitably equipped personal computer, as well as on the CD-I system.
CD-ROM XA will also add a new dimension to the type of applications
where CD technology can be applied, in both the professional as well as
the consumer markets. It will also form the bridge linking CD-ROM with
CD-Interactive.

Laser Vision

With the commercial introduction of LaserVision in the 1970s, many
pioneers in the computer industry saw the opportunity to combine the
audio-visual riches of videodisc with the processing power of the

personal computer. The result was interactive video (IV), a multi-media
hybrid that uses the videodisc as a computer peripheral.

In interactive video, computer software stored on a floppy or hard disc
generates text and graphics, and controls the access of sound and images
from the videodisc - in response to actions of the person in front of the
screen (who may be using a keyboard, mouse, touch-screen or other
device). Interactive LaserVision has developed as a specialist’s
technology, particularly effective for training and point-of-sales
materials. Despite this success, it has not penetrated into the broader
markets of education and the home to any significant degree, because of
the high cost of the hardware, and a lack of standards for both hardware
and software.

CD-I and Multi-media

Unlike the videodisc and CD-ROM, CD-I has been designed for
multi-media consumer applications. Its technological premise is the
complete integration of all media types. How CD-I achieves this, and the
full range of CD-I multi-media capabilities are fully explained in the
chapters that follow.

BEYOND MULTI-MEDIA: HYPER-MEDIA

The concept of hyper-media could have been invented to describe CD-I.
Its origins lie in the notion of hyper-text, a term coined by computer guru
Ted Nelson in 1965 as a method of linking related bodies of information
to allow the user to browse through different databases randomly. Owl
International’s Guide’ and Apple Computer’s ’HyperCard’ are two
recent software products that embody these notions. Creative software
designers are now using these same concepts to link different kinds of
media - hence the term hyper-media.

How does hyper-media work? An example can be taken from the Grolier
Multi-media Encyclopedia project, which uses a hyper-media approach
to integrate separate text, audio and visual databases. In this product, the
user can access the text of an article - say, the biography of Abraham
Lincoln - and use simple features of the system to find pictures of
Lincoln, orreadings of Lincoln’s famous speeches.

This article in turn may be linked to a "Time Machine’ database of
historical events, which might lead the curious users to, say, the history
of portraiture or photography and from there to a fully narrated
audio-visual essay, and so on.

Chapter 1: Introduction

6 CD-l: A Designer's Overview

All this must of course be defined by the design team - clearly, a massive
task when an encyclopedia is involved - but the concept is a relatively
simple one. With CD-I, you can link any type of audio, video and text
information on the disc to any other.

INTERACTIVITY - THE ’I’ IN CD-I

CD-I is founded on the concept of interactivity - that is, providing the
user with a means to interact with a program in a meaningful and
rewarding way. The success of the CD-I designer in developing
compelling interactive programs will ultimately dictate the success of
CD-I in the marketplace.

What is Interactivity ?

There are probably as many answers to this question as there are
interactive applications. Using an automatic bank-teller machine,
playing an arcade game, entering figures into a spreadsheet program,
accessing information in an on-line database - all these are examples of
interaction with a fully computer-controlled device. In CD-I terms,
interactivity is the method used to interact with the content of a program
which has been designed to respond in a very specific way to each
decision, choice or request made by the user.

The CD-I designer’s role is to balance the objectives of the application
with the degree of interactive control that is to be provided to the user.
This balance must be carefullly struck at all times - while a flight
simulator game might require continuous interaction from the user, an
instructional program or pop music disc might only require occasional
stopping and starting. (Remember, the user will purchase a CD-I disc for
its content, not to get a sore thumb pressing buttons!)

How to use Interactivity

There is no magic formula for developing good interactive programs,
although a full understanding of both the program’s content and the
user’s needs and desires is a prerequisite. The interactive video
experience can suggest some rules for the levels of challenge, reward,
review and so forth which can be useful for certain kinds of program.

However, in general, only one rule need apply: whatever the goal of the
program, make the journey to get there - the pathway through the content
- as interesting and compelling as possible. The TV screen directed by a
remote control device (a mouse, keypad or joystick) is the door into the
content. It is the designer’s task to help the user step through that door
and become immersed in a unique experience.

Interactivity: The Educator

Virtually all the experience for developing interactive materials (besides
video games) comes from the development of educational computer
software and interactive video. Educators have long known that
interaction with audio-visual learning materials enhances the
information transfer.

The success of interactive LaserVision training materials compared to
traditional training methods is still further evidence of the potential for
well-developed interactive programs.

Reports from training professionals indicate as much as a 40%increase
in retention can be gained over standard training techniques. In an
interactive program, the user is in firm control of the level and pace of
instruction, and actively engaged in the pursuit of knowledge.

However, the motivation of the student and the job trainee are quite
different from that of the average consumer, who is looking for
entertainment or information, but is not compelled to enjoy and benefit
from the experience. Consumers will not buy CD-I discs to get good
grades (although a CD-I SAT disc could prove the exception!).

Do Consumers want Interactive Programs?

Some observers have remarked that consumers do not want to interact
with their TV sets. Are they right? We do not yet know. Although a clear
case for interactivity can be made in the training and education markets,
nothing quite like CD-I has ever been offered consumers before, and no
body of research exists to supply us with an answer. The best market
feedback on which to base conjecture is the experience with video games
and home computers. Fora short period video games were phenominally
successful but, limited to ’shoot ’em up’ games with primitive graphics,
they proved a passing fad. The early home computers also failed to live
up to consumers’ expectations and quickly joined the video games at the
back of the closet.

Were these marketing disasters a fair test of the viability of interactivity
in the home? Hardly: it was too much to expect simple video games to
have any lasting value, and the software produced for the early home
computers was limited by the inadequate processing power and data
storage capabilities of the hardware, and the crudity of their visual
displays.

CD-I does not have these limitations. With its inherent multi-media
capabilities, massive storage capacity , and the powerful processors built

Chapter 1: Introduction 7

8 CD-I: A Designer's Overview

into the player, CD-I provides a system for software performance which
is well beyond existing home computer systems. With high quality sound
and video pictures, CD-I represents a new information resource,
delivering interactive programs on a wide range of topics and subject
areas. CD-I should appeal to the same variety of interests that drive the
sales of millions of books and magazines - certainly no previous
generation of software has had this capability.

There are categories of CD-I software that we cannot yet begin to
imagine, new combinations of information and entertainment waiting to
be invented. The fertile minds of the creative community will respond to
the challenge and opportunities of CD-I - the most powerful information
system ever known.

Chapter 2: The background of CD-|

CHAPTER 2: THE BACKGROUND OF CD-I

The last chapter introduced the key concepts behind CD-I technology -
interactivity and multi-media. This chapter presents an overview of
optical disc technology, and CD-I (Compact Disc-Interactive) in
particular. It traces the history of the medium and explains the different
kinds of Compact Disc (CD) now available - CD-DA, CD-ROM, CD-V
and CD-I. It also describes, in general terms, the technical and creative
aspects of CD-I discussed in later chapters, as well as the market
opportunities for this exciting new medium.

This chapter is essential reading both for those who intend to study this
guide thoroughly - designers, programmers and others preparing to make
CD-I software - and those who need only a conceptual overview -
presidents and vice presidents, administrators, marketing executives and
others not directly involved in the job of production.

THOSE CHILLY SATURDAYS

First, an imaginary look at CD-I as we may soon know it. Picture a
Saturday afternoon in the late autumn. Wintery clouds block the pale sun,
the air is crisp and cold. The chores are done for the day, (except to rake,
once again, the eternally falling leaves) and there’s nothing left to do for
a couple of hours. It’s a perfect time for sport. But, the summer sports
have finished, there’s nothing on TV until tonight, and outside - nothing
but fond recollections of warm days on the golf course.

TOT 8 Solna

[Archery [i]

Mcolf fA
[9 High Jump &! J
[9 Hockey

10 CD-I: A Designer's Overview

Chapter 6: Hot Shot
Sports for Design
Considerations

What a perfect time to try out that new edition of the CD-I golf game you
picked up at the supermarket this morning. A few preparations are in
order first - a snack, a soft drink, your favourite chair. Then, slap the
compact disc in the player, and you’re off to the sunshine again.

So where to go? Slide through a few of those good-looking choice frames
... Hot Shot Sport! Yes, that’s you - look out Jack Nicklaus, here you
come. You select golf from the list. Now choose a course. There are
several on the disc, but the main thing today is to find sun and heat, so
why not go for the big one ... that’s it ... the Augusta National, home of
the Masters. It’s a chance to challenge golf’s greats - and it’s always
warm in Georgia.

Before you tackle eighteen holes, why not sit back, finish off that hotdog
while watching a ’mini-movie’ about one of the most famous courses in
the world. That’ll get you in the mood, and besides you wouldn’t want
to smear mustard on the putter!

Because the CD-I system is both a computer and a television, you have
the best of both. The huge storage capacity of the disc offers a total
experience in action and pictures. On the one hand, you can play golf
with the interactive computer game: choose a whole round ora few holes.
Whatever control device you use - keypad, mouse or any other - you have
complete control over the animated figure on the screen, position, swing
speed and precise moment of impact between the club and the ball.

On the other hand, with CD-I, the computer-animated character is
combined with real photographic views of the fairway. The background
is not a crude computer-generated graphic, but a series of actual shots of
a world-famous course. CD-I also offers realistic high quality sound, to
heighten the sensation: if you want to imagine yourself in a tournament,
CD-I will even provide spectators who will follow your game with
interest and appreciation!

You don’t have to work at it. If you need a break from the stress of serious
tournament golf, youcan relax with some narrated video sequences about
the course. They describe the history, course layout, and some of the great
golf moments that have happened there. What a thrill to play a round on
one of the most famous courses in the world - from your armchair, CD-I
style.

The technology that can take a golf lover to Augusta or St. Andrews could
equally take an amateur archaeologist through the Pyramids or an
Etruscan tomb - or a gourmet cook into the great restaurants of the world

Chapter 2: The background of CD-I_ 11

for lessons from their master chefs - or a pop music fan into the studio
to create whole new videos to their favourite tracks.

All this is possible through a machine no larger or more demanding than
your VCR. CD-I truly combines entertainment with education,
information with recreation, in a form that the whole family can use and
enjoy for years to come. CD-I may seem like the technology of the future,
but all this is possible now - because of work that began over twenty
years ago.

ELECTRONIC PUBLISHING

Consumer Electronics Market

Traditional
Consumer
Electronics

Traditional
Publishing
Industry

Institutional Markets

The concept of publishing - that is, the distribution of information in
printed form - has altered little over the centuries, even if production
processes have changed greatly. However, the evolution of new
communication media - radio, film, television, audio and video
recording, as well as computing and telecommunications - has effected
changes which now present wholly new concepts.

The one feature these new media have in common is that they are stored
and delivered in forms that can only be read by a machine. They are
unintelligible to the naked eye. The concept has grown from punched
paper tapes and cards to magnetic tapes and discs, from factory
automation to telecommunications and mass data storage, from
entertainment through to every facet of information management. It is
now possible to exchange information instantaneously, worldwide.

12 CD-I: A Designer's Overview

Yet only recently has electronic publishing touched the world of print,
with the introduction of computer softwareand live on-line databases.
These have tended to be used for large volumes of information - typically,
financial, scientific or legal. It remained for optical digital discs to
introduce the concept to popular communications.

OPTICAL DISCS

Optical disc technology represents a great leap forward in the quality,
quantity and variety of data that can be stored,and what can be done with
that recorded information. Furthermore, optical discs are robust, virtually
impossible to pirate and so versatile that the technology has spawned an
enormous variety of consumer and specialist applications.

The original research program led to three product development projects.
The first of these involved data collection for company archival systems.
This resulted in the Megadoc system which was the forerunner of the
current Write Once Read Many (WORM) drives.

The other two development paths (leading to Compact Disc and
LaserVision) were for distribution systems. So disc encoding and
replication processes had to be developed as well.

The larger storage capacities and lower cost of production of optical discs
resulted in the Compact Disc - Audio (CD-DA) breakthrough. However,
optical media have been on the market since 1978 with LaserVision. This
video version of optical storage has found success in the educational
world in the form of interactive video discs. The digital compact audio
disc (CD-DA) was launched in 1982 and has proved highly successful
in the music listening world.

Optical media devices can be divided into two categories:

* Peripheral Devices: devices that can read the optical disc and transfer
the contents read to an external decoder.

Both LaserVision and CD-DA from inception were conceived as
systems.

Each of these types of device were originally intended for passive use.
The ’systems’ family was extended in March 1987 with CD-I and in July
1987 with CD-Video (CD-V is an amalgam of LaserVision and CD-DA
technologies, with a digital audio track added to the LaserVision disc).

Chapter 2: The background of CD-| 13

Optical Recording

Data Data

Capture Distribution

FUTURE

*) CD-ROM XA = CD-ROM extended Architecture

14 CD-I: A Designer's Overview

The original LaserVision and CD-DA specifications were also extended
to form the ’peripheral’ family branch with the introduction of CD-ROM
in June 1985 and LaserVision-ROM shortly afterwards.

Both peripheral and system variations are ’read only’ media, though only
CD-ROM and LaserVision-ROM are named as such. There is a great
deal of commonality among the LaserVision derivatives and among the
CD-DA derivatives such as the same basic disc mastering and replication
processes, and commonality of players for tracking and reading
disc-bound information.

The use of optical media, like that of many other types of product, ranges
from passive linear applications to fully interactive applications.

Existing media products like film and gramophone records could be
regarded as fully passive. Television with a remote control keypad has a
minimum level of interactivity requiring actions as simple as changing
the channel. The use of other types of product ranges from passive use
to full real-time interactivity in something like an automobile.

Optical read-only products span the same range. CD-DA for music
listening, and LaserVision discs for films are fully passive products. At
the other extreme are the interactive products of CD-ROM text
databases, the interactive videodisc programs used in education and
training, and the fully interactive consumer products of CD-Interactive.

All optical discs, whether intended for passive or interactive use, are
based on the same principles. Coded information from a master recording
on magnetic tape is burnt into a specially-coated glass master disc by a
powerful laser beam. The beam records digital signals as a pattern of
shallow pits and grooves on a long spiral path - rather like a conventional
audio LP, but much, much denser (and winding outwards rather than
inwards).

A series of copies taken from this glass master creates the metal stampers
which are used to press copies in a special plastic. A fine film of reflective
aluminium is laid over this, and sealed under a tough, clear plastic
topcoat: this allows the aluminium to shine through, but protects the data
well away from dangers of dirt, wear or rough handling. Optical discs
are virtually impervious to damage.

A low-powered laser in the reading head of the disc player bounces a
fine beam of light off the reflective surface of the disc through a network
of prisms and mirrors to a photodiode which decodes variations in
reflected light into audio and video signals, and computer data. The

Chapter 2: The background of CD-|

output of the player itself is compatible with existing domestic audio and
video equipment.

COMPACT DISC (CD)

The Compact Disc (CD) is one type of optical disc technology. It offers
a choice of formats for data of different kinds: CD-Digital Audio
(CD-DA) for top-quality sound recording, CD-Video (CD-V) for video
clips, CD-Read Only Memory (CD-ROM) for high-volume computer
data storage and retrieval, and CD-Interactive (CD-I) for the first fully
interactive combination of sound and pictures, computer text and
graphics on one system.

* CD-DA audio discs measure 12cm (4.75") across, andcan carry up to
72 minutes of top-quality digital audio per side.

® CD-V discs range from a 12cm ’single’ with six minutes’ video and
20 minutes’ audio to 20cm and 30cm (8" and 12") discs offering 20
minutes and one hour of video per side respectively.

* CD-ROM discs are essentially computer storage media, holding up to
600Mb of data on a 12cm disc - 150,000 pages’ worth of text infor-
mation and enough, say, for the complete white and yellow pages of
the whole of the East Coast of the United States.

CD-I will hold up to 650Mb on a 12cm disc, but can handle data from
a variety of source media, including natural video still frames (over
7800), audio (over 2 hours of top-quality sound or about 17 hours of
simple narration), text and graphics (up to 150,000 pages’ worth) or,
more typically, a combination of these under the control of the com-
puter program also stored on the disc.

COMPACT DISC-DIGITAL AUDIO (CD-DA)

Compact Disc-Digital Audio (CD-DA) - what most people now know as
*compact disc’ - is the most successful consumer product of the decade.
CD-DA was launched in 1982 and by the end of 1987 about 30 million
players and 450 million discs will have been delivered worldwide, with
current manufacturing capacity for over 100M discs a year. Consumers
embraced the new product because it offered very high quality sound
reproduction at an attractive price: as the next compact disc systems
emerge, offering multi-media and interactivity, the effects will be as
revolutionary as the emergence of sound and picture recording
themselves.

All CD technical specifications are based on CD-DA to ensure maximum
compatibility as the newer products come onto the market: the same

16 CD-I: A Designer's Overview

Chapter 3: National
Television Broadcast
Standards

plants can make discs for all formats, the new CD-V and CD-I machines
will play CD-DA discs.

ANALOG AND DIGITAL

Whereas other audio-visual media employ analog recording technology,
based on variations in electrical current, CD-DA, CD-ROM and CD-I
compact discs use digital techniques based on the more precise binary
computer code.

The striking difference between a compact audio disc and conventional
LPs and tapes illustrates one advantage of digital technology: the
digitally encoded compact disc is virtually free from the degradation or
noise’ inevitable in analog systems. This means not only crisper, cleaner
audio recordings, but also a medium reliable enough to carry computer
data as well as audio-visual signals.

INTERNATIONAL STANDARDS

To be compatible with existing television receivers and video monitors,
both videotape (VCR) and LaserVision are tied to national television
broadcast standards. This analog technology varies in both the
transmission of the basic video signal and its color-coding. A tape or
discmade for America, for example, cannot be played on a machine
purchased in Europe.

What distinguishes CD-I as an audio-visual medium is that, like CD-DA,
it is internationally compatible at its most basic level. As a result any disc
will work in any player, anywhere in the world. CD-I employs digital
technology. Wherever the system may be, a decoder within the player
adapts the video signal to the type of receiver or monitor to which it is
linked.

Furthermore, the standard was established by two leading manufacturers,
Philips and Sony, who have licensed over 150 other companies to make
discs and players.

Technically, this makes a CD-I disc as universal as a book or audio
recording. However, CD-I has another feature: one disc can hold enough
data to present the same material in several languages. Sound, pictures,
text, graphics and computer programming are stored separately, and only
mixed within the player during the actual presentation.

The disc’s large and flexible storage capacity can thus handle the text,

Chapter 2: The background of CD-|

pictures and sound to present a substantial program in a choice of national
languages.

COMPACT DISC-INTERACTIVE (CD-I)

Natural Evolution
1985
Yellow
Complete Specification

Physical
Format
Spec.

Multi Media
(Consumer +
Institutional)
Full Compat.

Music Peripheral
(C.E.) (Professional)
Full Compat. Partial Compat.

The CD-I Full Functional Specification, or Green Book, was issued in
March 1987. Amongst a number of ’logical requirements’, the following
stand out:

Compatibility with the CD-DA specifications (the Red Book), so that
a CD-I player can handle all CD-DA discs; (some CD-ROM and some
more recent CD-DA players can also be upgraded to CD-I).

Compatibility with existing consumer electronic products so the new
CD-I player can not only send sound through a home stereo, but also
pictures as well as sound through the home TV.

The entire interactive multi-media presentation must be contained on
one CD-I disc for playback on a CD-I player - unlike interactive La-
serVison, there need be no extra computer and software.

Future proof’ technology that takes account of new and proposed
standards, and allows room for later enhancements.

THE CD-I PLAYER

The CD-I specification defines the minimum standard or base case. This
may be an integrated CD-I system or a separate decoder (or *black box’)
to upgrade an existing CD-DA player. The basic drive is identical in all

18 CD-I: A Designer's Overview

Appendix C: CD-RTOS

compact disc players and includes control circuitry for the laser read-out
head.

At the heart of the CD-I player is its operating system. This controls the
rest of the system. CD-RTOS, Compact Disc - Real-Time Operating
System, was developed to presentinteractive multi-media in ’real-time’
- that is, in direct response to the user in front of the screen.

The CD-RTOS, held in the player’s memory handles the basic
management of the system, including the synchronization of audio,
video, text, graphics and computer data pouring in from various sources
for decoding and co-ordination into the final presentation on the screen.

On all compact discs, information is organized in tracks and sectors.
Typically, on a CD-I disc all the material in one application is held in
one track, which comprises many small sectors of individual audio, video
and text. These sectors are played through the system at 75 per second.
What appears on the screen, or is heard through the loud speaker, depends
therefore on what these sectors contain. If they are full of audio, there
will be no new pictures arriving at the screen; if they are full of video
pictures, there can be no sound. So the effective CD-I producer must
learn to balance the creative requirements of his program with the rate
at which sectors can be played back - the amount of space available in
the data channel. It is therefore essential for the producer to come to an
early understanding both of the features available for program making
and what space in the data channel each individual feature takes up.

CD-I AUDIO

CD-I offers a choice of quality levels for both audio and video: higher
levels produce higher quality, but take up more space in the memory
banks and in the data channels which send information from the disc to
the player. The choice of quality level is vital since CD-I must typically
handle a variety of data at any one time.

CD-I offers six audio options - three sound quality levels in either mono
or stereo: A-Level (mono and stereo) is equivalent to the first play of a
brand new high quality audio LP; B-Level (mono and stereo) is
equivalent to the very best FM radio broadcasts, transmitted and received
under optimum conditions; and C-Level (mono and stereo) is equivalent
to AM radio transmitted and received under optimum conditions.

16 channels each of 72 minutes’ duration are available to play back the
audio. A-level stereo uses 8 of them at once, so, allowing for some
computer control data as well, a CD-I disc can contain just over 2 hours

Chapter 2: The background of CD-I

A-level stereo if there is nothing else on the disc; on the other hand, a
disc full of C-level mono (using one channel at a time) and containing
nothing else could produce a talking book over 16 hours long - or one
program in 16 different languages.

In the critical trade-off between space and quality, A-Level stereo is best
reserved for a musical interlude, while C-Level is used for commentary
or background music when other data need a good share of the available
resources.

CD-I VIDEO

CD-I also offers a choice of video picture quality: normal resolution for
most video pictures; double resolution, for better definition of computer
text and graphics; and high resolution, to meet future standards in digital
television.

CD-I defines four main types of full-colour picture, and a choice of
techniques to record and process these as economically as possible.

*Natural’ images such as photographic still frames use a video-based
technique called DYUV, which offers very subtle and realistic color
and shading.

Computer text and graphics can use RGB (Red, Green, Blue) compu-
ter coding, but more often use a Color Look-up Table (CLUT), which
creates a palate of up to 256 colors at a time.

Simple cartoon style drawings use Run-length coding for large blocks
of single colour. These simple images can be processed quickly
enough to create animation on the screen.

Where conventional video creates and records special effects when the
master tape is edited, CD-I can simulate many effects from still pictures
recorded on the disc. These include:

* cuts, wipes, fades and dissolves between images;

® scrolling horizontally or vertically;

© partial updates which change the picture in only part of the screen.
These can be achieved through the use of CD-I’s four visual planes: a
small ’cursor’ plane at the front of the screen, two full-screen planes and

a background plane for a fixed backdrop. A variety of images can be
created by building up a composite picture using two or more of these

20 CD-I: A Designer's Overview

Chapter 4: The Design
Brief

Chapter 5: Designing for
Production

Chapter 6: Hot Shot
Sports

Chapter 6: French
Phrasebook

planes simultaneously. Clearly, there are considerable opportunities for
the creative designer.

THE DESIGN PROCESS

In the early stages at least, it is likely that most CD-I projects will be
based on successful work in other media, where the familiarity of the
concept or title will attract consumers unfamiliar with the new
technology.

Work may well begin with documents such as the client’s brief, the
design company’s proposal, and the treatment and contract documents
then worked out between the two.

Essential production documents will likely include a storyboard, a
flowchart, and scripts for both narration and text screens. The storyboard
illustrates both the appearance of the finished presentation and the
interactivity mapped out in the flowchart. In CD-I as in interactive video,
creative energy is concentrated at this stage, and not inproduction, for
the scale and complexity of the job means that every detail must be agreed
before the functional work of creating and assembling the component
parts begins.

Many things may happen in parallel during the assembly stage: shooting
original pictures, creating graphics and text screens, recording
soundtracks, preparing computer programming and so on.

When all the material is brought together, it is encoded and compressed
to create data files for testing on a simulator (a large computer with hard
disc storage). Here it must be rigorously tested and evaluated, and
adjustments made until all the interactive elements work smoothly
together. Only then is it prepared for digital encoding and pressing as a
compact disc.

Even then, the work of implementation and evaluation may continue for
months or years, particularly in so new a medium, to learn from the users’
responses what people really want and expect from this exciting new
medium.

This chapter has given an overview of the family of optical recording
technologies and a short description of the salient features of CD-I. The
following chapters will build on the brief look into CD-I that was
presented here to include descriptions of the full range of media options
and the stages in the process of designing a CD-I title.

Chapter 3: What CD-I cando 21

CHAPTER 3: WHAT CD-I CAN DO

The last chapter provided a broad overview of the background to the
development of optical disc technologies and briefly described the
features and capabilities of CD-I, This chapter looks at the wide ranging
media palette available to the CD-I designer. Specific reference is made
to the audio-visual concepts behind CD-I technology. More details of
how CD-I provides these capabilities, in technical terms, are found in
Chapter 7.

USING AN INTERACTIVE TELEVISION SET

The heart of CD-I lies in its name - interactivity: CD-I’s ability to search
and locate the information requested by the user as whenever it is
required. Also, CD-I can present this information in combinations of
photographs, cartoons, music, speech and text, all of which can be called
up at will by the user.

CD-Iis aimed primarily at consumers. It will also be extremely attractive
to the professional and electronic publishing markets for such topics as
education and training, catalog shopping and travel information.

CD-I has the great advantage of being able to do many of the things that
other optical disc products can do, but at the same time offering many
other features as well. It will both play super hi-fi music from CD-Audio
discs and also display the massive amounts of text and graphics typically
stored on CD-ROM discs. In addition, CD-I has been designed to meet
consumers’ expectations of high quality video in still and moving
pictures, photographs, cartoons and computer graphics.

The enormous repertoire apparent to the consumer is backed up by
technical features which the designer must master in order to take full
advantage of the opportunities presented by CD-I.

CD-I will play back in all current television standards and, through its
interactive features, allow the user to pick and choose, mix and match.
It also offers an enormous range of special effects and technical features.
In addition to those commonly available in existing audio visual media
like a choice of mono or stereo sound, video wipes and fades, and 3-D
graphic images, CD-I also offers, in real-time, a choice of still or motion
video, three different audio levels, three degrees of picture resolution,
four coding techniques, and four separate image planes with and without
transparency.

22 CD-I: A Designer's Overview

The description of these features presented in this chapter should give
the designer sufficient background to sketch out the concept of a CD-I
project. More detailed technical information is available in later chapters.

AN EXAMPLE: CD-I GOLF

The golf game described in the opening of the previous chapter makes
use of the wide ranging options of a CD-I system.

A game of armchair golf combines the thrill and expertise of an
arcade-style game with the high quality sound and visual images we
expect from television.

The control device attached to the system - a specially designed remote
control unit, a mouse, a joystick or perhaps a full computer keyboard on
some - allows interaction with the animated character. Moving the cursor
to the golfer’s hat, allows the whole body to move on the screen, aligning
the direction of the shot. Moving the cursor to the ball, controls the swing.
Holding down a switch or button starts the backswing, releasing the
switch releases the foreward swing and tapping once again determines
the moment of impact with the ball. The resulting flight of the ball has
been determined by the skill of the player.

With CD-I, you don’t simply jump into a swing-and-hit game. CD-I golf
is a multi-media experience. Earlier, we chose the Augusta National
course in Georgia. Before we started play, we finished the hotdog and
coke, while watching a documentary about the course.

Chapter 3: What CD-l cando 23

It was no different than watching a program on television. It showed us
top quality photographic views of the course, aerial views of the hole
layouts, even illustrated maps of the course with inset video sequences
showing the great comeback victory of Jack Nicklaus at the Master’s in
1986. And the sound - marvellous! Narration, background music, even
sound effects that were almost as pure and clear as listening to music
without pictures from a CD audio disc which we could put on the same
player.

Many of the effects that make CD-I programs a delight to watch and
listen to, or even control and interact with, are created within the CD-I
system itself. Whether you are planning a golf game, devising a tour
through a museum, designing a language learning course or a music
survey disc, understanding and using these features opens up a vast range
of program possibilities for the CD-I designer.

AUDIO

CD-I audio can be played back through a home hi-fi system as well as
through a domestic television set. It therefore not only meets existing
hi-fi audio expectations, but also offers hi-fi audio in combination with
high quality video images.

Quality Levels and Capacity

CD-Digital Audio (CD-DA) can store just over an hour’s worth of stereo
sound of the highest quality on a single 12cm disc - but it devotes the
whole of the compact disc’s storage and processing capacity to this end.

Whilst CD-I can of course play back CD-DA quality audio, space has to
be made available, both on the disc and in the processing channels, for
the other media - natural video images, graphics and text. The system
offers the designer three quality levels of audio in mono and stereo.

The higher the quality level, the more data has to be stored and, at the
appropriate moment, transferred from the disc. Therefore, the choice of
quality level can be critical for some titles. It must be determined by the
disc designer according the needs of his design, the disc capacity, and
the space available in the data channel.

On many occasions, there will have to be a trade-off of quality level
against transfer rate or disc capacity. It is therefore important that the
designer understands from the outset the difference among the quality
levels.

24 CD-I: A Designer's Overview

These audio levels are:

Data Channel Occupancy and
Maximum Playing Time

DA A B Cc
(Super Hi Fi) (Hi Fi) (Mid Fi) (Voice)

4 hours 8 hours

Max playing time: 1 hour

MONO

ec
Max. playing time: 4 hours 8 hours 16 hours

¢ A-Level audio, equivalent to the first play of a brand-new high-quali-
ty audio LP, but, because it is read by a laser beam, without any of the
noise commonly created by contact between needle and disc. Stereo
audio of this quality requires only half as much information as CD-
DA, and thus occupies only 50% of the data channel, leaving the re-
maining 50% available to handle other material (video, graphics and
text). Mono A-level audio occupies only 25% of the data channel.

If the whole of the disc were to be filled with A-Level audio and nothing
else, just over two hours’ pure stereo (or four hours of mono), could be
recorded on a single disc.

© B-Level audio, equivalent to the very best stereo FM radio broadcasts,
transmitted and received under the very best conditions. Stereo audio
of this quality requires only 25% of the data channel and so leaves the
remaining 75% of the data channel available for other material.

If the whole disc were to be filled with B-Level audio, over four hours’
stereo or eight hours’ mono music could be stored.

® C-Level audio is equivalent to AM radio when broadcast and recei-
ved under optimum conditions. Audio of this quality in mono occu-
pies only 6% of the data channel, leaving 94% for other material.

If the whole disc were to be filled with C-Level mono audio, over 16
hours audio could be played back.

Chapter 3: What CD-l cando 25

Where CD-DA handles a single channel providing up to 72 minutes of
stereo sound, CD-I offers up to 16 channels in mono. Sound can be played
continuously as music or narration.

In addition sound can be transferred from the disc to the player and stored
as soundmaps in the CD-I system short-term memory to respond to
interactive cues from the user.

The wide range of potential applications for CD-I will require a variety
of audio effects from high-quality stereo music to mono voice-overs and
narrations. The contribution made by the audio track at any one moment
will help determine what quality level (and so, disc and channel space)
it merits: this may vary from segment to segment within the program as
attention shifts between audio, video, text screens and pure interaction.

In a music video, it may be worth spending up to half the available disc
resources for high-quality sound; in a multi-lingual production, the
multiple parallel tracks can be used for example to provide over an hour’s
narration in 16 different languages - and allow the user to switch between
these at will.

Soundmaps

In addition to taking the sound directly from the disc, short sound
sequences can be stored in the player’s own temporary memory for ready
access and processing without further reference to the disc. These
sequences are known within the CD-I specification as ’soundmaps’.

Typically, a soundmap might be used where a short sound effect may be Chapter 6: Hot Shot
repeated quite often during the course of a segment or program, e.g. in Sports: Audio

response to a user’s interaction. Once the soundmap is stored in the

decoder’s RAM memory, itcan be called up at any time, while the player Chapter 5: Basic

itself retrieves information of other kinds from other areas of the disc. Principles: System RAM

Soundmaps may be mono or stereo and at any quality level. Two mono
soundmaps may be mixed to achieve one stereo soundmap, or sound data
from the disc may be mixed with a soundmap to produce another
soundmap altogether. Soundmaps may be used once or looped back as
often as they are needed.

Soundmaps may also be used to improve the performance of a CD-I
program by providing temporary storage. Sound can be pre-loaded from
the disc at a convenient moment into one soundmap into order to release
the whole of the data channel at a later moment for other material; or to
be mixed and played back with the output from another part of the disc.

26 CD-I|: A Designer's Overview

Specialist Use

In addition to playback of stored sound, the CD-I system microprocessor
can be used to generate its own sound. Specialist titles such as computer
music applications could come provided with optional hardware such as
a synthesizer keyboard to be attached to the system.

Audio Control

The stereo audio output levels may be controlled in a number of ways
including, for example, panning from right to left, or attenuating under
the control of the application on the disc.

VIDEO

CD-I has been designed to meet consumers’ expectations of the highest
quality video in still and moving pictures, photographs and computer
graphics. CD-I shares many features with other audio-visual media -
particularly, specialeffects such as cuts, wipes, dissolves and so forth.
However, whereas in conventional video production, these effects can
be created and recorded only in the editing suite, CD-I can also create
effects within the player itself.

It also has a unique combination of additional video features to offer the
designer - playback on all international television standards, a choice of
image resolutions, as well as access to four video planes.

National TV Broadcast Standards

CD-I will play back on all television broadcast standards. Currently,
there are two main (incompatible) broadcast television standards used
variously throughout the world, each with different screen display sizes.
The system used in North America and Japan employs a 525-line screen
updated at 30 times per second. This is known as NTSC. The PAL system
is used in Britain, most of Europe, Australia, Africa and South America
is based on a 625-line screen. A full picture is updated 25 times per
second.

CD-I overcomes this problem of incompatibility. A disc which has been
made to the CD-I standard for international use could be bought in New
York and played on a machine in London, Leningrad or Bombay - just
as for CD music discs.

Resolution
CD-I provides three levels of video resolution just as it offers three levels
of audio quality:

Chapter 3: What CD-l cando 27

Normal resolution is equivalent to the best picture quality obtainable
on a normal broadcast television receiver; it is likely that most CD-I
images will appear in normal resolution.

Double resolution is equivalent to the best picture quality obtainable
on a standard computer color monitor, and provides better reproduc-
tion of high-quality computer text and graphics.

High resolution is equivalent to the best quality digital picture gene-
rated in the studio to AES/EBU standards.

These resolutions can best be expressed in terms of the number of picture
elements (known as pixels) which appear across the screen both
horizontally and vertically.

NORMAL RES. DOUBLE RES. HIGH RES.

Safety Area

The video picture tends to drift beyond the edge of the visible screen,
particularly in older sets. This is known as overscan. Because of this,
both television systems define a safety area in normal resolution 320
pixels across, which is 210 rows high in 525-line systems (NTSC), and
250 rows high in 625-line systems (PAL/SECAM). In broadcasting this

28 CD-I: A Designer's Overview

is known as the television safe title zone, and is virtually guaranteed to
display adequately on all television receivers. When designing for
international compatibility the CD-I designer should keep important
information within adisplay area of 210 rows of 320 pixels (normal
resolution).

Screen Resolution in Pixels

NTSC PAL/SECAM
360 384

Normal 240 280
720 768
Double 240 a 280
Nee
720 768
High 480 560
Compression

CD-I employs compression techniques to store and retrieve audio data
as efficiently as possible. A variety of techniques are also used in picture
coding to minimize the sometimes considerable demands of high-quality
visuals.

Where conventionally-coded RGB graphics (Red, Green, Blue - the
standard computer graphic coding system) might occupy 300k bytes of
digital storage space and take nearly two seconds to load, compression
techniques reduce this to around 100k bytes, loadable in less than a
second.

Chapter 3: What CD-Icando 29

Coding Techniques

Four coding techniques are available to the designer. One, known as
DYUV, is best suited to photographic images. Two, RGB 5:5:5 and
CLUT are more appropriate for text and complex graphics. While,
finally, Run-length coding is available for text, cartoons and graphics
which use large blocks of color.

Natural photographic images are best handled through a compression Chapter 5: Disc Capacity
technique known as DYUV, or Differential (or Delta) YUV coding. This

is based on conventional broadcast television and video technology,

where Y represents the luminance of a video signal, and U and V its color.

A combination of coding techniques based on compressing the color

signals, and coding only differences between consecutive pixels results

in a compression ratio of 3:1 in comparison with conventional RGB

coding.

When subtle shades and gradual changes in tone and texture arerequired, Chapter 7: DYUV Images
such as in photographic images, DYUV is ideal. It is not suitable for text
and those graphics which call for a crisper and less subtle display.

In cases where high resolution is needed for a natural image, a technique
known as QHY (Quantized High-resolution Y) can enhance the
luminance (’Y”’) signal to achieve this (the color resolution is sufficient
already). Instead of direct high resolution DYUV coding, which would
require four times the disc and memory space, high resolution is achieved
by interpolation. QHY is used to correct the interpolated values where
these differ significantly from the true values. This feature is not in the
Base Case hardware specification.

Three coding methods are available for graphics: RGB 5:5:5, CLUT and
Run-length coding.

RGB 5:5:5 is available only in normal resolution, and provides a Chapter 7: RGB 5:5:5
compression ratio of 1.5:1 - that is, around 200k bytes per full screen Images

image - by reducing the number of RGB levels from 256 to 32 (or 8-bits

to 5-bits) per RGB. This still leaves a range of 32,768 colors to choose

from. It is very good for graphics, and best suited to user-manipulated

graphics such as paint or drawing applications. Because other coding

techniques are available which compress the image at a higher ratio with

little loss apprarent to the user, RGB 5:5:5 is a relatively inefficient

means of coding.

CLUT, stands for Color Look-Up Table. This is the locationin memory Chapter 7: CLUT Images
where a set of colors that the designer has defined is stored. CLUT is
available at both normal and double resolution. Compression - at about
the same 3:1 ratio as DYUV, and double that of RGB 5:5:5 - is achieved

30 CD-I: A Designer's Overview

Chapter 7: Run-length
Images

by restricting the total number of colors available for any given image to
a range of 256 or fewer, pre-selected by the designer from some 16
million ultimately available. The contents of the CLUT can be defined
as having 256 levels each for red, green and blue, so the total number of
colors available to the designer for a single image is 256 x 256 x 256 =
over 16 million. The total range of colors available is greater than for
RGB 5:5:5, but the number of colors on the display at any one time is
limited. Since the human eye can only discern some 5,000 to 10,000
colors in any single image, this restriction in colors only becomes a
problem for highly accurate representations of natural images where
subtle gradation is essential.

When used at double resolution only 16 colors are available instead of
up to 256 in normal resolution.

~—

Clut Length of Run Clut Length of Run
Location (number of Pixels) Location (number of Pixels)

Run-length coding is very economical for certain types of graphic image
suchas cartoons which use large blocks of color. Not only are the number
of different colors limited, but also the number of color changes on any
one line. Images like this can be stored, retrieved and manipulated on the
screen very efficiently through Run-length coding.

Run-length coding uses the CLUT to define the colors to be used. The
color choice per image is limited to 128 at normal resulution and only
eight colors at double resolution.

The big advantage of Run-length coding is that a typical cartoon or line
drawing will require only 10k to 20k bytes of data.

Chapter 3: What CD-Il cando 31

It is usual to store text in a compressed form using the character coding
techniques of computers. However text can also be treated as an RGB,
CLUT or Run-length coded graphic image. The CD-I specification
defines a standard character set covering all Latin alphabet languages,
including a number of special characters for currency symbols, diacritic
marks, and so forth. Alternative character sets and fonts can be created
transferring information from the disc into the decoder, so CD-I is truly
multi-national and multi-lingual.

Image Planes
Chapter 3: Visual Effects

Chapter 5: Screen Effects

Direction
of view

32 CD-I: A Designer's Overview

The. picture which the viewer sees on the screen can be composed of
several image planes which appear one behind the other. CD-I offers up
to 4 image planes. The first is a limited area 16 pixel x 16 pixel single
color cursor plane. Behind this can be one or two full-screen image
planes,which may be coded and displayed individually or together in any
combination of DYUV and CLUT.

Alternatively these two planes can be merged into one plane so that an
RGB 5:5:5 image, which uses twice as much data as the other coding
methods, may be displayed.

Behind these planes is a fourth or background plane (not in use in the
illustration) which acts as a backdrop in cases where all or part of both
image planes are transparent. The backdrop may be a single color
selected from a fixed range of 16. In some CD-I players, the backdrop
may be replaced by an external video source.

The experienced audio visual designer will readily appreciate the
enormous potential offered by this combination of planes.

Motion

The capabilities of CD-I have so far been described in terms of still
images. CD-I is nota still image medium and the designer is encouraged
to use motion where it is needed. However, because of technical
considerations, care must be taken in designing motion material. It is
important for the designer to understand the ways in which motion is
achieved.

The first technical consideration is disc capacity. In computing terms,
650 Mb of data is an enormous store. In feature movie terms, however,
one would be forgiven for believing that 7,800 DYUV video images will
get eaten up in about 4.5 minutes. But of course a movie does not show
7,800 totally different images in 4.5 minutes - a lingering romantic movie
might show only two or three basic images in that time with quite subtle
differences between most frames. So disc capacity can only be calculated
with precision when the actual information load - the quantity of separate
bits of information to be shown on the screen - is known. This quantity
can only be measured after the coding techniques and frame rate are
known. This is another example of how important it is that the designer
understand the technical requirements of CD-I - that means in the first
instance reading right through Chapter 7 and the Appendices. This
chapter is not the place to delve into so much detail. Suffice it to say at
this stage that the designer should be aware that disc capacity is a

Chapter 3: What CD-I cando 33

constraint, but not a great constraint, if CD-I’s features are to be used to
the full.

There are a number of ways in which images can be given motion when
this is required.

Full motion on the screen can be achieved through partial updates - that
is, a technique which changes the picture in only part of the screen at one
time. The rate at which data can be transferred from the disc means that
only about 13% of the screen can be updated fast enough to achieve full
motion video.

However, cognitive full motion - that is where no subjective motion jitter
or blur can be seen by 95% of the population - requires a minimum of 10
frames per second. The designer can achieve moving images at this rate
covering up to 50% of the screen by using software coding techniques.

If an even larger picture area is required, full motion video can be
displayed by using chroma key - a video technique frequently used on
television in which one image plane of moving elements within the
picture is electronically keyed over a still or scrolling image in the second
plane. The size of image which can be achieve on CD-I depends, as
always, on the amount of new information to be transferred from the disc.

A similar multi-plane technique is an integral feature of traditional cel
animation in which the still elements of a scene are constantly re-used
and a separate cel for the parts which move is laid over the top - typically
at the rate of 10 to 15 frames per second. Run-length coding is a good
technique to use in achieving full screen full motion animation since the
typical Run-length image over the full screen may occupy between 8k
and 15k bytes.

Again, the update rate will, of course, depend on the actual complexity
of the images and the amount of space available for them, which in turn
depends on the audio quality level and other data needed at the time.

Limited animation effects may be produced by re-defining the CLUT Chapter 6: Pop
colors. A singalong sequence in a music application might show a ball Showcase: Screen
moving across the screen to help the listener follow the words ofasong, Propertion

To simulate the movement of the ball as it follows the text, a series of

circular shapes are laid out on the screen, which can be colored using

locations in the color look-up table.

The balls can be made transparent by making each one the same color as
the background. By successively changing the colors to yellow and back

34 CD-I: A Designer's Overview

to the background color, the ball appears to move across the screen. The
CLUT contents are successively re-defined as the ball moves across the
screen.

This simple technique, of course, may require a number of spare CLUT
locations, but it can be very powerful.

Another possibility is dynamic CLUT update. By re-defining the CLUT
from line to line down the display, the number of colors available in an
image may be expanded to the limit that can be loaded into one field -
about 2,000 colors.

By re-defining CLUT colors in these ways, it is possible to create a range
of highly dynamic effects.

VISUAL EFFECTS

CD-Ican not only store and retrieve virtually any image as a high-quality
still or graphic, but the range of special effects available within the
system itself includes cuts, wipes, dissolves, granulation, scrolling and
animation -enough features to rival most things to be found in modern
video editing suites !

Single Plane Operations

Some effects can be achieved on a single image plane, but many require
both. Single plane operations include cuts, sub-screens, scrolling, mosaic
effects and fading.

Cuts

The cut - the sudden change from one image to another - is the simplest
visual effect. In a single plane this means cutting between images stored
in the player’s own temporary memory. A number of images may be held
this way and used at a rate faster than any designer would likely want to
use them.

Cuts can also be performed, of course, by switching directly from one
plane to another.

Partial updates are rapid cuts in a single plane covering only part of the
full screen image. Rapid cuts give full motion video.

Chapter 3: What CD-l cando 35

Sub-screens

7-bit
CLUT

SUBSCREEN
Sbit
CLUT

Each image plane may be divided into a number of horizontal bands ‘Chapter 6: Interactive
called sub-screens which can, if necessary, use a different coding Under Fives
method. So images of different kinds can be shown together without

recourse to both image planes. In the example shown in the diagram, the

screen is divided in three parts: the main picture can be two DYUV

images and the band in the lower portion of the screen a text image using

double resolution CLUT or Run-length coding. The number of

sub-screens, their positions and their boundaries formed by horizontal

lines may be defined by the designer at will - in the extreme, each of the

525 or 625 lines could be a sub-screen. The advantage of using this

technique in the instance illustrated (a pattern recognition game for under

5s) is that it frees the other visual plane for a larger, more visible cursor.

Scrolling

36 CD-I: A Designer's Overview

Images may be scrolled horizontally or vertically. Simple examples
include the vertical scroll of a picture of a clocktower from the base
towards the top, or a horizontal scroll across an image stored in memory,
perhaps a panoramic view much wider than the screen itself. In either
case, the screen acts as a window onto the larger image. The relative
positions of the image and the screen can be changed to facilitate smooth
scrolling at any speed.

A combination of sub-screens and scrolling allows a central area to be
scrolled between two fixed areas of the screen at top and bottom.

Mosaic Effects

Mosaic effects can be used for granulation and magnification of an
image; basically, they involve reducing the resolution of the image
through one of two mechanisms called pixel hold and pixel repeat. Partial
updates of reduced resolution images covering a larger part of the screen
can be achieved in this way.

Pixel hold retains the whole picture but reduces the resolution by making
the image appear granulated. This is achieved by taking a pixel value and
holding onto it for a defined number of pixel positions both horizontally
and vertically. The different value (or colour) of the other pixels in the
original image is ignored. This technique can be used with any image
coding method, including DY UV and Run-length; the hold factor can be
any number from 1 to 255, which may be independently set for horizontal
and vertical directions.

Pixel hold is used for granulation effects where the size of an image
remains constant, but a blocking effect is produced as shown in the figure.
In this example, with a hold factor of two in each direction, every other
pixel in both horizontal and vertical directions is expanded to four pixels
on the screen, although the total image size remains the same. By
changing the hold factor, the resolution of an image can gradually be
reduced until it becomes unrecognizable, at which point, the image can
be cut to another and the hold factor gradually reduced so that the new
image now appears.

Pixel repeat magnifies a portion of the image without providing greater
detail. Each pixel is displayed a number of times in sequence, as shown
in the figure. A pixel can only be repeated 2, 4, 8 or 16 times. This
technique can only be used with CLUT and RGB 5:5:5 coding so pixel
repeat can be used for reduced resolution images, which require less data
from the disc, or to magnify or zoom images.

Chapter 3: What CD-l cando 37

38 CD-I:A Designer's Overview

Chapter 6: French
Phrasebook

In the diagram, a part of one image is magnified by a factor of two both
vertically and horizontally, so each pixel in the original becomes four in
the displayed image. Unlike pixel hold, only a part of the total image can
be displayed. Partial updates of reduced resolution images covering a
larger part of the screen can be achieved in this way.

Fade

The brightness or intensity of any image can be varied from black to full
intensity through a range of 64 levels. Different parts of an image can be
given different levels of brightness than other parts, so only parts of an
image fade. It is more likely that this effect will be used to move from
one image to the next in combination with two-plane effects.

Two-Plane Effects

All of the coding methods except RGB 5:5:5 allow for two separate
image planes. The use of two planes, together with transparency, can
provide a range of very useful effects including transparency in parts of
an image, mixing of images, dissolves and wipes.

Transparency

The most obvious reason for providing transparency is so that the second
image plane, or part of it, can be viewed through the first image plane.
For example, a cartoon image in the first plane can be placed over a fixed
scenic background in the second plane.

There are three methods of achieving transparency.

Chroma Key

Chapter 3: What CD-l cando 39

The first is the use of chroma key or color key, a technique which has
been used in the video industry for many years. Red, Green and Blue,
the RGB colors, for each pixel are compared to the color key value, which
is defined in 8 bits of red, green and blue.

If the pixel color and the key color match, then that pixel is made
transparent revealing the background. Since the color key function
operates on the final RGB values, it is independent of the coding method.

Mattes

OPAQUE
[__] TRANSPARENT

Dyer ha aa e

B :
9

%
i 2€
3S Cm

6 Shout shout let it allout

-/ #><C IP J a

A second method for achieving transparency is the use of mattes. A matte
is an area of any shape - regular or irregular - which is defined on the
screen so that the area inside the matte is transparent and the area outside
is not.

This is achieved independently of color key or any other transparency
mechanism. A number of mattes can appear on the screen at one time
and two mattes can overlap.

Transparent Pixels

In RGB 5:5:5 only, there is a transparency bit available for each pixel,
so any combination of pixels can be made transparent.

Since RGB 5:5:5 does not allow the use of two planes, only the backdrop
or external video can be seen when the transparency bit is used.

40 CD-I: A Designer's Overview

Dissolves

Before

After

As explained, it is possible to adjust the intensity of an image. By mixing
the two image planes together at different intensities then translucency
can be achieved. One example of this is the dissolve, where one image
can turn into another by changing the intensity for each image so that
one increases as the other decreases, and so one apprear to dissolve into
the other.

The use of mattes allows mixing and dissolving to be restricted to parts
of the screen defined by those mattes.

Wipes

Chapter 3: What CD-lcando 41

This is another way to change from one image to another which requires
two planes. For example the figure shows a horizontal wipe, where one
image changes into another by a wipe from from right to left. This can
be achieved quite simply through the use of mattes where a simple
rectangular matte, for example, starting from the right hand side of the
screen and gradually moving across to the left hand side produces a wipe.
Wipes can of course be horizontal or vertical and in either direction.
Mattes can, of course, have different shapes and be varied in size, so for
example a diamond shaped matte can start very small and gradually
increase in size until the new image covers the full screen.

REAL-TIME INTERACTIVITY

This chapter has so far discussed the audio-visual elements in CD-I, but
another essential feature is of courseinteractivity. Digital data on CD-I
discs may contain audio, video, text or graphics. Data can also be used
to control the presentation itself and to interact with the user in front of
the screen. All this must happen in ’real-time’ (rather than, say the
artificially fast or slow speeds common in computing), which makes
special demands on CD-I technology.

Real-time Data

Audio and video are of course played from the disc in real-time. In
addition, data of any kind from any part of the disc can be accessed - that
is, found and retrieved - at random. This access is instantaneous if the
new data is close by, and no longer than two seconds if the laser has to
travel from one extreme position to another.

CD-I decoders can handle a variety of tasks in parallel and in real-time:
this is essential if audio, video, text and graphics are to be synchronized
to reach the user in the right order and at the right time.

Synchronization

Synchronization is very critical in the kind of complex multi-media
presentation that CD-I offers. In the example shown here, the visual of
the golf club swing must synchronize with a carefully-timed audio effect
as the club hits the ball. The timing of the synchronization cue depends
upon the instant that the player - acting through a remote control or other
input device - determines, which in turn designates the exact moment of
contact of the club on the ball.

42 CD-Il: A Designer's Overview

Chapter 5:
Microprocessor

Chapter 5:
Synchronizing to Video

This sound effect can be handled as a sound map pre-stored in RAM and
played through the audio output channels at the required moment. The
basic images are taken from disc and stored in RAM as drawmaps, and
the animation sequences are handled by manipulation of the appropriate
drawmaps.

Fairway

G
Graphic Character Animation

Synchronization is controlled through data recorded on disc, and may
occur in response to pre-recorded ’ triggers’ in the program, or the user’s
spontaneous response to the program. These elements of the technology
are explained more fully in Chapter 7.

USER INTERFACES

Interaction would be impossible without suitable user interfaces - that is,
devices which allow the user to control or react to information on the
screen by making choices, decisions and requests. An important element
of the CD-I design task lies in the preparation of these essential user
interfaces.

Physical Interface

The CD-I system specifies an X-Y pointing device as the main user
interface. This could be a keypad, a mouse, a joystick, a light pen, a
graphics tablet. It is always equipped with two trigger buttons which may
be used in a variety of ways determined by the designer.

Chapter 3: What CD-l cando 43

The specific device chosen may depend on the application or range of | Appendix C: The
applications used, and individual CD-I players may well allowachoice. Pointing Device

For example, a mouse is suitable when sitting at a table close up to the

screen for making selections or doing simple drawings on the screen. An Chapter 6: Interactive
infra red keypad would be most appropriate for those listening to and —_ Under Fives
watching a disc in television mode - from the other side of the room.

As an option, an alphanumeric keyboard may be provided with CD-I
players. This will be necessary where the user is required to input long
or complex textual responses.

For players which do not have a QWERTY keyboard, a simplified
keyboard can be displayed on the screen and the X-Y pointer used to
select characters from it.

Interacting with the User

A pointing device may be used to move the cursor around the screen. By
controlling the cursor position, the user can select menu items or buttons,
move level controls and so forth. Scrolling around an image which is
larger than the display screen may be achieved through directional
arrows or sliders on the screen.

The screen may therefore be designed to include certain areas, such as
buttons - also known as hot spots or action areas. These allow the user
to make selections, go back to a main menu or pause within an
application. They may be displayed explicitly on the screen so that the
user can see them, or the designer of a game for example may have
decided that they should be invisible.

Interaction with the user is an essential part of a CD-I system and the
designer must take great care to ensure that the display on the screen is
always clear and designed in such a way as to make it inviting. The user
must always understand readilty what to do - and, indeed, what can be
done at any point in the programme.

CONCLUSION

This chapter has shown the range of features available to the CD-I
designer. The combinations into which these may be woven may at times
be bewildering, and certainly make the design process much more
complex than that for any other audio visual medium. However, rich
rewards await the imaginative designer who can master the techniques
and who can conceive and develop programs that will set the world new
standards of creative achievement.

Chapter 4: The Design Brief 45

CHAPTER 4: THE DESIGN BRIEF

Chapter 3 surveyed the media palette available to the CD-I designer.
This chapter looks at the special creative and technical challenges posed
to the designer by CD-I, and suggests some pertinent questions for the
would-be producer to consider. It outlines typical stages in the design
and production of a CD-I project, and discusses the skills and
backgrounds required of members of the design team.

The previous chapter showed the wide variety of sound and picture
making effects incorporated into the CD-I system. It is evident that a high
level of planning and control is necessary if truly exciting and
intellectually rewarding programs are to be created.

Like all successful publications of high quality, CD-I discs must be
designed on the basis of a very clear understanding of the user and
particularly what the user wants or expects from the program and the
technology. So the very first task for the designer to address is not how
to exploit the exciting opportunities that the technology offers but, rather,
the age old issues of who will want the product, how it will fit into the
market place, and what attributes of the subject matter make it suitable
for the medium.

The successful designer is one who can answer these questions
accurately, and as a result develop a product that will interest and excite
the audience. CD audio has established the compact disc as a quality
product in the minds of the public. CD-I must build on that reputation
and success. It is therefore important that the first discs meet consumers’
expectations of a product which is new, and unlike anything else on the
market: informative, entertaining, worth the money, and yet not too
frighteningly modern or complex - in fact, somehow familiar.

DEVELOPING THE DESIGN BRIEF
To develop a clear design brief, the potential CD-I designer will need
first to address several key questions:

* How does this title fit in with the others already on the market? Does
it duplicate, complement or supercede existing products?

® Who will want to buy the disc? - especially if they have to buy a CD-
I player, too

46 CD-I: A Designer's Overview

DESIGN PROCESS

Chapter 4: The Design Brief 47

* How will the interactivity be used? How will the program retain its in-
terest over many playing sessions? What will happen if the program
is left unattended or the user is unable to respond?

As CD-I grows, the industry will be able to address these and other
questions in more detail. However, the experience of other media -
including interactive video - offers some guidance.

Why choose this title? Is it best suited to the market opportunities of
CD-I?

CD-Ican combine high quality photographic images, full-color graphics,
animation, text and sound in a range of qualities and presented in new
and potentially exciting ways - and it can store huge volumes of this
information, compactly in digital form on a single small disc: this we
know.

The CD-I designer must decide whether any of these features adds real
value to a product which could be made in another, more conventional,
medium. The CD-I title must use CD-I’s unique attributes appropriately
- and be readily accessible to the consumer. The temptation to exploit
the technology simply because it’s there, or seek novelty for its own sake
must be firmly resisted!

Particularly in the early years, adaptations from other successful mass
market publications - books, video or computer software - will be
important. People unsure of the technology will respond to a familiar title
or concept, and this will form a bridge from the new product to
established markets. However, to capture these new sales, it is important
that CD-I really does add new dimensions to a product already successful
in another medium. Otherwise, why buy the CD-I version?

The creation of a totally original concept for CD-I will present the
greatest challenge to the CD-I designer - but the reward will be a
purpose-built product which makes full use of the CD-I’s unique
potential, in creatively imaginative ways.

A series of titles which make use of use the same basic design and
production processes will help to diffuse the work and cost of developing
software for the new medium. Typically these might include games,
instructional materials, reference works and entertainment such as pop
music programs or children’s shows.

48 CD-I: A Designer's Overview

Ask:
® Is this acommercial, promotable, consumer product?
© What does CD-I technology contribute?

© If it is an adaptation from another medium, what is the audience pro-
file and sales history of that medium? How does that compare to the
immediate prospects for CD-I, outlined below?

How does this title fit in with the others already on the market? Does it
duplicate, complement or supercede existing products?

The first CD-I designers will waste effort if they try to produce a title
which is already in production elsewhere, or does not fit into the range
of products already indevelopment. There is simply not the room for two
similar titles in such a new market.

Content providers, those who own titles already successful in other
media, will look to CD-I as a means of re-exploiting the market. Fears
that the new medium will destroy the market for the old product are
ill-founded.

It has been established, for example, that the film of the book and the
book of the film stimulate demand for each other.

The Grolier Multi-media Encyclopedia - one of many titles currently in
production - is not a more conveniently shaped and packaged clone of
the 20-volume print Academic American Encyclopedia, but an entirely
different and complementary product providing education, information
and entertainment in ways not previously envisaged.

Ask:

* DoI know enough about other titles that are being produced? Where
can I learn more - from the publishing and AV trade press? From pro-
fessional associations and groups? From the grapevine?

Who will want to buy the disc? - especially if they have to buy a CD-I
player, too?

Chapter 4: The Design Brief 49

They are consumers who:
® like electronics for home entertainment (stereo, video, computers);
* ownaCD Audio player;

® like to be the first on the block’ with any new product or status sym-
bol.

Each potential CD-I project must be thoroughly analysed for its
commercial potential in that market. Likely applications will include
education and information, but entertainment is expected to lead the
market.

The true multi-media nature of CD-I will allow designers to draw on the
very best of published material and titles will not be limited to the types
of electronic material currently available for home computers and in
arcades.

Typical applications might include armchair travel and language courses
with authentic sounds and pictures supported by masses of text; music
tuition or a pop video that contains information and pictures as well as
top-quality digital soundtracks, or even the Japanese ’karaoke’ option to
switch off selected voices or instruments and supply the track yourself!

Chapter 6: Typical CD-I
Applications

50 CD-I: A Designer's Overview

Chapter 6: The Grolier
Multi-media
Encyclopedia

Chapter 6: Country
House Murders

Simple special effects and control commands could allow users to create
original programming from the sound, pictures and text on the disc.

The initial target audience appreciates a product with educational value.
CD-I can offer a constructive alternative to commercial television, and
even something which encourages people of different ages to play or
study together.

CD-I should prove a popular medium with children, with entertainment
that is engaging, interactive and informative.

Ask:

© What is the demographic breakdown of potential purchasers? How
does this profile compare with that of the first CD-I market?

How will interactivity be used?

The core of any CD-I title is its interactivity: it is the elegance - or
otherwise - of this design that makes or breaks the program.

The user will have to be given the maximum encouragement to interact
with CD-I. It must be clear at all times what is to be done next; there must
never be an occasion when the user can get lost within the program; the
user must always feel in control. Of course the designer may spring a few
surprises - one would expect as much in a game or similar application -
but this must not be so sharp as to disorientate the user.

?U TAMKHAMEN

AINE LF ELGUPT

Chapter 4: The Design Brief

CD-I must also maintain the delicate balance between passive reception
and true interaction. It is tiring to interact all the time - especially in
programs which fully exploit the length and complexity permitted by
CD-I’s storage capacity and multiplicity of effects. Interactivity must be
perceived as true added value, and not merely a gimmick which becomes
annoying when the novelty wears off.

Equally, the interactions themselves must be substantial: trivial rewards
could put the new user off not only that program, but the whole concept
of interactive home entertainment.

A key element in this design is the input device. A CD-I base case system
can support an infra red keypad - a familiar device to owners of remote
control television receivers - a tracker ball, a mouse, or a joystick which
may be more familiar to owners of home computers. The choice for any
particular application will depend on what room the CD-I player is in
and how far the viewer is from the screen, but is crucial in establishing a
friendly interactive environment. Similarly, menu design and other
visual aids and prompts will also play a critical role; most computer users
are familiar with the technique, but others will need encouragement and
guidance. For example, menus can be hidden until they are needed. They
must certainly be honed until they are as clear and straightforward as
possible.

Action regions supported, for example by InVision, can be defined so
that whole parts of the screen can be used to select routes through the
disc: point to the door and give it a ’push’ and you’re inside. A whole
language of icons must be devised to guide and prompt users through a

52 CD-I: A Designer's Overview

range of actions and responses. Pictograms and visual aids are especially
useful in multi-lingual applications, and can streamline the appearance
of menus and choice frames.

Audio prompting is a very immediate and friendly way to encourage
interactivity. Technophiles can be frustrated by the slowness of audio
prompting, but it has worked well for general audiences in LaserVision
applications, particularly in point-of-sale projects.

Paradoxically, part of the elegance of an interactive application may lie
in allowing the viewer to watch some segments passively. Effective use
can be made of this ’Auto-play’ mode to sell the application, and to
intrigue the user into trying new alternatives. It can also help the user to
improve his performance by showing how the application can best be
run. We do not yet know how consumers will react to interactivity at this
level: what models we have - video games, for instance - are not always
directly comparable with typical CD-I applications such as multi-media
encyclopedias and programs to teach reading.

Ask:

¢ What is the most appropriate way of making interaction easy and fun?
° Is the user interface fit for its purpose?

DEVELOPING THE IDEA MAP

Having established the appropriateness of the title, medium and level of
interactivity to the market and to the audience, the next task is for the
designer to give shape to these ideas.

The design process may begin with a letter of agreement from the
publisher or other ’content provider’. If the client originally submitted a
brief for tender, and the production company replied with a proposal,
there may now be a contract and a treatment agreed between the two.

The basic document which opens the door to pre-production will
typically address the following fundamental aspects of the project.

¢ The Treatment,

Chapter 4: The Design Brief 53

DESIGN PROCESS

” DESIGN | “BUDGET _
il TEAM ~~ ' oF be

ae
ag ea

. _\ / PRINCIPL lll a

_ DESIGN. \~
ICOMPLETED

54 CD-I: A Designer's Overview

TREATMENT

The starting point for production is the treatment. It is a narrative outline
describing the proposed content or storyline, often written from the point
of view of the user, including a description of how the interaction will
appear to the user.

At this stage the precise design of the content has not been developed,
so the treatment would present a hypothetical scenario, offering a
preview of the possible program scope. It enables a publisher as well as
the design team to be certain that the initial project brief has been
understood with regard to what will eventually be produced, mastered
and bought by customers.

Ata later stage of the design process this treatment may be redeveloped
in a more detailed form, when the storyboard phase begins and when a
clearer picture of the budget has been outlined. A detailed treatment is
an essential document to brief scriptwriters and a potential source of
publicity for the application.

THE DESIGN TEAM

Programmers

Graphic Content
Designers Providers

CD-I demands a wide variety of design and production skills, some
drawn from established media such as print and electronic publishing,
computer programming and audio-visual production, others emerging as
the technology itself develops.

Chapter 4: The Design Brief

Yet what CD-I demands perhaps more than any other medium is not only
close co-operation within the design team, but also a sound
understanding by everyone involved of both the technology and the
individual disciplines which make up a CD-I program. Each must
understand the technology to make the best possible use of its resources;
each must also understand the potential (and limits) of disciplines other
than their own, to choose the best medium for each message.

In CD-I, after all, the choice between video still, computer graphic, text
screen and audio is more than a creative one, when storage space and
speed of transmission are at issue. The way in which the user interacts
with the program is both a technical and a creative decision which affects
every member of the team from the software specialist to the screen
designer and the scriptwriter. For a CD-I project to be truly interactive,
the whole design team must appreciate the range of technical and creative
elements involved at every stage.

Thus, in CD-I, familiar job titles like ’producer’, ’scriptwriter’, ’ graphic
designer’ or programmer’ may cover arange of new and complex skills.
Perhaps the early years of CD-I will offer talented and adventurous
people the kind of opportunities that the early years of film-making and
television did.

A Typical Design Team

° a project producer/director who will exercise creative control and
ensure all materials are correctly integrated;

° a CD-I author who will control the overall interactive design and
may write the script and the text screens as well;

° a designer of audio-visual material and software;

° a graphic designer to originate images and to design user interface
screens;

To compete on the consumer market, CD-I programs will have to offer
at least as much as established favourites in television, video, hi-fi and
games - which means not just excellent sound and picture quality, but
also significant entertainment value and ease of use for an audience not
accustomed to high levels of interaction with recorded media.

Accordingly, those with a background in film, TV or corporate video
will have much to contribute to CD-I - and some new ideas to absorb and
existing attitudes to change, in the shift from linear to interactive media.
Yet CD-I is in many ways like a book, and the traditional skills of writers

Chapter 5: Basic
Principles

and

The Mechanics of CD-I
Design

56 CD-I: A Designer's Overview

and editors as well as book designers, illustrators and art directors can
profitably be transferred to CD-I. The publisher’s role is similar in both
media.

BUDGET AND SCHEDULE

A target production budget must be developed at the very start. Although
it will be far too early to develop this in any detail, block sums can be
assigned to the major elements of production. These should include the
costs of the interactive design and storyboards, image, sound and text
origination, This will be constantly reviewed during the initial
production phase until the completion of the storyboards, when a detailed
and firm budget can be fixed.

The CD-I designer must have a clear idea of the cost of the project. This
might seem to be difficult to calculate, at least until a few titles of
different kinds have been produced. However, many parameters exist in
traditional audio visual production to enable realistic target estimates for
all but the final pre-mastering phases to be developed.

However, a thorough understanding of the technology is essential to
create a design which can be realized with the money available. For
example, the use of CD-I’s multiple soundtracks and image planes means
that a single disc can be designed for a wide international market, but the
additional cost of translating all the text and recording all the audio tracks
must not be forgotten.

Nonetheless, once the parameters of the content, budget and the
production schedule have been fixed, many elements of the project are
as predictable as those of any electronic video program: salary costs can
be calculated once it has been decided how many people will be involved
and for how long, material resources from office space to technical
facilities are predictable, as are the replication costs of the discs
themselves - all familiar elements of any audio visual or publishing
project to which firm estimates can be fixed.

There has currently been so little experience in CD-I production that it
is not useful at this stage to assign outline costs to production. Many of
the early projects being developed have budgets in the range $250,000
or more, while adaptions of existing products will cost much less. Large
prestige productions developed from scratch will cost a great deal more.
It seems reasonable to expect typical budgets to be around $250,000,
somewhat higher than a single volume book, but much lower than low
budget movies.

Chapter 4: The Design Brief

It is worth remembering that successful international book publishing
ventures have made good profits on budgets well below and sometimes
well in excess of these sums. Yet the potential returns from sales in a
truly international market place (once it has fully developed) are likely
to match the most successful books and movies.

In addition to the budget, a schedule of work, together with information
on how that work is to be monitored and achieved has to be agreed.

Again, it is too early to be helpful about the length of design and
production time. Simple conversions of existing interactive LaserVision
products which did not seek to exploit CD-I’s unique features could be
completed in weeks, while a high original, bespoke product - an intricate
tour of one of the world’s great art galleries for example - might take
years.

Whatever its length, the schedule must show, at specific critical points,
where and how the review process will take place - who will be involved,
and what issues are to be addressed at any particular time. Review periods
may occur at fixed intervals (monthly or weekly, for example) or at the
end of pre-determined production phases.

The initial project documents at the treatment stage are likely to set the
terms for critical stages in the evaluation - say, completion of
storyboards, during simulation, on delivery to the client, and after the
program has begun to be used by its intended audience under real
working conditions. In such a new medium, though, it is entirely possible
that these criteria may be revised as the project develops.

The Production Process

In order to estimate the schedule of work and ultimately a projection of
the cost of production, it is essential to have a clear sense of the phases
that a typical application would go through as it passes from the idea
stage through the design stages to eventual realization as a program on
compact disc.

PHASE ONE: is the Idea Map - the development of a treatment, budget
and schedule. It will also identify members of the design team and the
probable outside content sources, facilities and studios that may be used
in the production phase.

PHASE TWO: is concerned with the development of prototype
storyboards and flowcharts. These will define the range of problems that

58 CD-I: A Designer's Overview

Chapter 6: French
Phrasebook

Chapter 6: Hot Shot
Sports

need to be solved before production can begin and typically consist of
archetypal frames, stored in electronic form or as hard copy.

Prototype storyboards will show:
° screen design issues
° elements of the user interface
° a list of resources for sound and pictures
° an outline of coding proposals and other authoring issues.

The prototype flowchart will show:
© the overall design for the interactive application, in both micro
and macro terms
° any design issues which may affect the preparation of applica-
tion software.

PHASE THREE: At this stage, problems are solved and definitive
designs take shape. Electronic storyboards will present prototype frames,
using master material for both sound and pictures. This material will later
be encoded to the CD-I specification. The user interface will be
comprehensively defined and the interactive flowchart will be finalized.

PHASE FOUR: Electronic storyboards can now simulate parts of the
program. And the design and production parameters can be tested, using
the most appropriate hardware.

PHASE FIVE: The shooting script will now contain a comprehensive
list of audio-visual requirements, including how these are to be arranged.
All the picture requirements will be listed according to their original
format, and their encoding mode, and all audio elements will of course
also be included.

A detailed budget and production schedule with reviews and income and
expenditure schedules should be ready by this stage. A database
management system will help to monitor all the individual production
elements comprehensively throughout the project.

PHASE SIX: Production proper can now begin. After the video data has
been ’captured’ (whether assembled from existing material or freshly
shot, or a combination of the two), it must be processed and integrated.
The video images are composed and reviewed, the predetermined coding
is applied, and design personnel review the data which is then entered
into the database monitoring system.

Chapter 4: The Design Brief

PHASE SEVEN: The audio data is composed, edited, and mixed down,
and the appropriate sound levels are selected. The audio material is then
reviewed in the same manner as the video data.

PHASE EIGHT: Finally, authoring - the creation of the control code - is
undertaken when the design is fully implemented. All the audio-visual
components are linked together with all text and user inputs.

The project is now ready to be reviewed and evaluated.

CONCLUSION

The development of the Idea Map provides a broad picture of the CD-I
project into the design and production phases that stretch ahead. It sets
up the control procedures that will help to maintain the continuity of the
work that may be carried out on several fronts at the same time. With a
well laid out document, the actual process of designing a CD-I
application can begin with confidence that it can and will be realized.

The mechanical design tools essential to the coordination of a CD-I
application will be covered in the next chapter. These include the nature
of the CD-I authoring environment, some basic principles of data
management unique to CD-I, and the way in which various effects can
be combined.

Chapter 5: Designing for Production 61

CHAPTER 5: DESIGNING FOR PRODUCTION

Chapter 4 looked at the early stages of the design process, in which a
general plan for the development of a CD-I project is outlined. This
chapter will look into the next stage, in which the design enters a more
detailed phase - that of storyboard development. The chapter will outline
the proposed systems that will make up the authoring environment and
looks at some basic principles of CD-I design and the way in which
elements are brought together in the CD-I system to produce a program
that attracts and holds a user.

DEVELOPING THE STORYBOARD

Once the first stage of the design process, dealt with in the previous
chapter - the treatment, budget and schedule, and the design team - have
been approved, the project moves into a detailed design stage. The task
here is to develop one or more levels of storyboard that prepare the way
for the production and authoring of program material into CD-I digital
data. Applications in Chapter 6 show the types of storyboard style that
may be used depending upon the stage in the production process. The
techniques described here are a typical way to carry this out. However,
different design studios may carry out the same types of task in different
styles.

The storyboard page in the French Phrasebook application would beused = Chapter 6: French
prior to the final authoring stage, when the precise nature of audio levels, | Phrasebook
screen coding requirements and interactive branching must be known.

The example illustrated for the Hot Shot golf game uses a more Chapter 6: Hot SHot
generalized descriptive form to convey an overview of necessary Sports
information about screen layout, video planes, audio quality,

synchronization and the possible transitions to subsequent sequences.

These could change before they reach the final authoring stage.

Storyboards would usually be developed alongside a flow diagram that Chapter 6: French
lays out the branching patterns at each stage of a sequence. It is essential Phrasebook

to know the pathways between stills or linear sequences, in order to

estimate processor requirements, images and sound sequences that must

be available for user choice, and the way in which users can escape from

a section or get assistance if confusion arises.

THE CD-I AUTHORING ENVIRONMENT

Before examining the detailed aspects of design itself, it is worth
outlining the parts of the CD-I authoring environment that will become

62 CD-I: A Designer's Overview

DESIGN PROCESS

e _\ / PRINCIPLES | [| .

Chapter 5: Designing for Production 63

available to assist designers with the potentially complex tasks of
working within data management threshholds for capacity and transfer
rates. Some of these tools will be useful at the beginning of the design
process. They will help the designer to simulate the general concepts of
the program in rough form or as further briefing documents for graphic
artists or live action directors. Other types of tool will be restricted to the
final software authoring stages.

It is important to be aware that, at the time of writing, a variety of | Appendix C: CD-RTOS
hardware and software design, production and testing tools are under 2d Invision
development. A number of manufacturers and software houses have

indicated delivery of protoype authoring systems and software during

the second and third quarters of 1988. What follows is a description of

the principles being followed rather than a precise specification of any

particular hardware system or set of software tools.

The tools, collectively known as the authoring environment, will
normally consist of a modular system, expandable from a single,
low-cost workstation to a network of stations linked together into a
complete production facility. It will be a distributed system, to allow
teams of creative people to work simultaneously on an application, with
easy access to common data and control information.

The authoring system will assist the designer in the writing of application
code, the capturing of audio and video, and the editing or digitizing into
CD-I formats, of these design elements. The edited data will be stored in
the system’s powerful database, combined with control elements into
disc files. Testing, another function of the authoring system, will involve
playing back assembled files ona simulator. It will consist of a read-write
storage medium and special hardware which will emulate a CD-I player.
The simulator will ensure that the application is tested in a real-time
interactive setting.

The Designer’s Station

The designer’s station will commonly be a single, low cost workstation
with CD-I simulation capability used by a design team for storyboarding,
scripting, program development and testing. It can be used
independently, by attaching a local disc and/or tape streamer to the
station. Audio, video and other data will be transferred in digital format
to and from the station via tape.

The Production Facility

The complete production facility will consist of several workstations and
dedicated audio and video servers. The facility will also use a studio
server, which will consist of a number of large read/write discs, write

64 CD-I: A Designer's Overview

once optical discs for archival storage, and magnetic tape for output to
the mastering plant.

Disc Building

The authoring system will be used for design, production and integration
of an application through each stage of the disc building process. During
the scripting phase, the designer needs to be sure that the elements of the
application will not exceed the limitations of the CD-I player. For this
purpose, an on-line Constraint Analyzer is being developed which can
check the script against the known boundaries ofthe player. Disc band
width, memory limitations, seek times and other key design
considerations will be analysed using this software.

When design parameters have been set, the authoring system will guide
the assembly process, which consists of data acquisition, presentation
editing and disc building. Data acquisition involves the capturing,
editing, and encoding of audio and video. Audio acquisition can be
handled outside the CD-I authoring environment, as many studio
facilities are capable of handling the complete digitization process.

Presentation editing links audio and video with programming.
Simulation will be controlled by the presentation editing software, which
will also provide data-specific editing tools, and tools for the building
and testing of real-time records.

_ Lastly, disc building combines the control information with the encoded
audio, video, text and application programming. These elements are
structured hierarchically into records and files, and processed into forms
needed by the master tape generator.

Standard Data Formats

The data used for CD-I applications will come from many sources. It will
need to be translated into standard data formats to ensure compatability
and quick, accurate data transfer. The standard formats include reading
and writing routines and a standard library of access routines.

Scripting Subsystem

The term ’script’ is defined here as a set of structure, display and control
commands which are used by the programmer in creating CD-RTOS
modules. Simulators will read the script as the CD-I control and
command language. The scripting subsystem will allow the designer to
describe the program structure, controls and logical flow.

Chapter 5: Designing for Production 65

Audio Subsystem

The audio subsystem will accept digital audio, and convert it to any of
the allowable CD-I formats. This will be accomplished by the application
of digital signal processing algorithms. Digital to analog conversion will
occur by the connection of a D/A converter to the output stages of a
digital audio processor. Non-real-time synthesized sound will be
produced by generating PCM files within the authoring system.

Video Subsystem

Images acquired from standard video sources will be filtered and
processed into digital images. Standard image processing techniques
such as noise reduction and color balancing will be supported. Manual
editing and enhancement of images will be accomplished using digital
paint systems. Once an image has been digitized, filtered and edited, it
will be encoded into one of the accepted CD-I image formats such as
RGB, CLUT or DYUV. Previewing the processed images will occur
using a CD-I simulator.

The Presentation Editor

The Presentation Editor will assist in the creation and integration of
real-time records. It will also link data into blocks and display these
blocks for testing purposes. Both functions will use a simulator and
CD-RTOS modules to support communication between the simulator
and host system.

The creation of real-time records is one of the most important aspects of
CD-I development. A real-time record editor will be involved in
synchronization techniques, interfacing with graphics routines, and the
generation of correct file interleave factors.

The Database Subsystem

The central problem in the creation of CD-I discs is the management of
the massive amount of data which a disc will contain. The amount and
complexity of data can be overwhelming, as it combines scripts,
storyboards, CD-RTOS modules and source files, as well as multiple
versions and types of images and sounds. Copyrights for all of these
would belong to different people and organizations.

The database manager will provide a central control and management
facility for coherent access to CD-I data. It will parcel out the data to the
various utility programs and provide data locking to prevent two
programs from modifying the same piece of data. As modifications are
made, the database system will maintain a history of such changes.
Access to data will be controlled, to provide data security and privacy.

66 CD-I: A Designer's Overview

Testing and Simulator Subsystem

The simulator will provide quick turn-around testing of ideas during the
creation process. It will also test the final disc image before mastering
takes place. In the basic authoring environment, the simulator will be a
compiled’ facility. In the future, further development in the area of
simulators may enable them to provide interpretive facilities. These
would interpret database files to build the CD-I datastream, allowing
faster interaction than the compilation technique, which requires each
disc image to be rebuilt after every change.

BASIC PRINCIPLES

The authoring tools dealt with in the first section of this chapter will
relieve the design team of some of the potentially burdensome
calculations required to keep track of data in the CD-I system.
Essentially, a designer wants to know what will happen on the screen or
speakers of a television system and what will happen when a user
participates in the various elements of interactive program material.

This is a major creative task. Software tools that can simulate the effects
both at the general level and later at the specific authoring level, as well
as provide feedback on data management, will release the design team
to apply energy to solving creative problems rather than mathematical
ones.

CD-I is not a form of video, like a VHS cassette player, but a
computer-based technology: sound and pictures are simply aspects of a
digital databank. That data can be transformed into video pictures, stereo
sound, text, or it can remain in the digital domain to guide the direction
of an interactive application.

The key to understanding CD-I and interactive design is to think about
it as a computer process, rather than a television or print process. The
design tasks can then be seen in terms of gathering and organizing
different types of data that are made available to a CD-I user through the
interactive computer interface.

While the CD-I designer need not immediately understand all the
complexities of the technology, it is important to grasp the basic
principles of CD-I design - the rate at which data can be transferred from
the disc, the amount of data required for the various elements that make
up aCD-I program, and the memory space available both on the disc and
within the system. For example, how much space is required to store each
type of data? How fast and to what locations can data be moved before

Chapter 5: Designing for Production 67

decoding and playback? What proportion of the available memory
locations and data channels are consumed by each type of CD-I effect?
What will happen on the screen while data is being located at a new place
on the disc? How many tasks can be handled by the main processor?

In this section some of the elementary calculations needed to keep track
of storage capacity and processing power are introduced. Design for
CD-I requires a regular monitoring of both of these by estimating the
data amounts involved in creating desired effects. Eventually they will
be handled by authoring stations. Nevertheless, areasonable fluency with
these numbers will give a greater depth of understanding about the types
of effect that can be incorporated into CD-I progams. Appendix A gives
a detailed account of the precise figures for storage quantities and data
rates.

Basic Principles
© Tracks and Sectors

* Disc Capacity
© Data Transfer Channels
° System RAM

® Main Processor Power

Tracks and Sectors

In any CD-I disc, data is stored on a track of sequentially-recorded Chapter 7: CD-I Sectors
sectors. A sector contains approximately 2k bytes of data. The precise

amount differs depending on whether it is Form 1 (usually audio and

video program material) or Form 2 (usually text and program control

data).

Chapter 7 deals with this subject in detail. What is important to remember
here is that the CD-I player reads data at a constant rate of precisely 75
sectors/second, no matter where on the disc that data is located. This is
equivalent to approximately 170k bytes per second (that is, 75 sectors
multiplied by the precise sector size - which can vary slightly). So one
sector (just over 2k bytes of program material) can be changed and
interleaved in the stream of data coming off the disc at a rate of 75 times
every second

A sequence of program sectors can be grouped together to produce a
program module. The key to interactive program design is to break
sequences into short modules that enable the user to couple them together
through choices made at the screen interface.

68 CD-I: A Designer's Overview

Appendix A: Technical
Specification Summary

Disc Capacity

The structure of the CD-I disc is explained in detail in Chapter 7, but
essentially, the usable total storage space on a CD-I disc is 650 megabytes
(Mb) of digital data - enough space for roughly 150,000 pages of text on
a single disc.

Audio capacity is calculated according to the quality level used. One
second of Level A stereo uses 85k bytes; Level B stereo 42.5k bytes; and
Level C stereo 21.3k bytes. Mono at each level requires half these
amounts. So if 650 Mb of disc space is available for sound, simple
mathematical calculations can be made to determine how much of any
particular audio level a CD-I disc may store.

Unfortunately, a megabyte is 1,048,576 bytes, not 1,000,000 bytes.
Those not blessed with an understanding of computers must accept that
there are 1024 bytes ina kilobyte and 1024 kilobytes in a megabyte (1024
x 1024 = 1,048,576).

So, to return to our task, if 650 Mb of disc space is available only for
sound, then just over two hours’ of Level A stereo can be stored (650 x
1024 = 665,600k bytes divided by 85 = 7830 seconds = 130 minutes =
2 hours 10 minutes), or a prodigious 17.5 hours of Level C mono, but of
course with no other data to accompany it.

Single visual images occupy various amounts of disc space depending
on how much of the screen each fills, the screen resolution (normal, high
or double), and what coding technique has been used.

It should be remembered that an image covering only part of the screen
requires only a percentage of the space of a full screen image. For
example, a picture occupying half the width and half the height of the
screen would take up 25% of the screen (not 50% as the mathematically
unwary might suppose).

A single full screen DYUV image in normal resolution and occupying a
full 8-bit plane fills 360 pixels x 240 pixels in NTSC (384 x 280 in PAL).
Each pixel needs 1 byte, so the mathematical calculation is quite
straightforward: 360 x 240 = 86,400 x 1 byte = 86,400 bytes per full
screen natural DYUV image in NTSC (107,520 bytes in PAL). The
86,400 bytes (107,250 bytes in PAL) can be rendered into kilobytes by
dividing by 1024. The answer is 84.38k bytes (104.74k bytes in PAL) -
call it 85 and 105k bytes. So, in simplistic terms, if 650 Mb of disc space
is available for natural DYUV images, at least 7,800 different full screen
DYUV images can be stored on and replayed from a CD-I disc (650 Mb

Chapter 5: Designing for Production 69

x 1024 = 665,600 kilobytes divided by 85 = 7,830). If only 25% of the
screen is used, a CD-I disc can store over 30,000 separate DYUV images.

A single RGB image occupies both 8-bit planes and can be used in double
resolution only. It therefore requires 170k bytes NTSC (210k bytes PAL)
of storage space.

CLUT graphics occupy varying amounts of space depending upon the
type and coding method. 8-bit and 7-bit CLUT, like DYUV, occupy full
screen planes and require 85k bytes (105k bytes in PAL) of storage. 4-bit
and 3-bit require only 1/2 byte per pixel or 42.5k bytes of disc space.

Using Run-length coding for 7-bit and 3-bit CLUT extends the capacity
of a disc considerably. For example, a run of 10 pixels of one color - say,
blue sky - would take only 2 bytes, one to indicate the color and one to
note the length of run. The full economy becomes apparent in, say, a
large expanse of a single color running for 50 lines at 360 pixels/line.

The maximum run of a single line is 255 pixels, however, setting the
Run-length to 0 sets the distance to run to the end of the line. Each line
would therefore need only 2 bytes of storage instead of the 180 bytes of
a normally-coded 4-bit CLUT graphic. So 50 lines of one color would
need 50 x 2 = 100 bytes instead of 50 x 180 = 9,000 bytes divided by
1024 = 9k bytes.

Of course, the actual requirement of any image depends on the data
content of that particular image. With experience, it should be possible
to estimate disc capacity fairly accurately. These concepts are explained
more fully in Chapter 7.

Data Transfer Channels

The CD-I player reads one sector on the disc at a time and allocates it to
a data transfer channel. There are up to 16 possible data channels
available for audio information. Each other data type can have up to 32
channels.

Channels are a useful metaphor for the way in whichdifferent kinds of
data must be interleaved into the data stream. For example, a television
receiver may intercept all the available broadcast channels, although
most TV sets can display only one at a time. In the same way, all the
sectors in the data flow are picked up by the player and allocated to their
respective channels, but only those called upon by the user during that
particular run of the program will be used. (Although in CD-I, if not in
most TV sets, more than one channel may be played at the same time.)

70 CD-I: A Designer's Overview

Chapter 7: CD-I
Decoder: Audio
Processor

Language Learning

ENGLISH VO. _

VIDEO
CHANNEL

[_] German voice over in USE.

In the diagram, voice-overs in three languages are interwoven with
background music and photographs. During the sequence, the user could
decide to change from English commentary to German without
interrupting the flow of music or pictures, since all three voice tracks are
running in the data stream simultancously.

Data Transfer Channels: Audio Playback

All audio in the CD-I standard is digitally encoded until it is finally
played back through analog amplification and speaker systems. This
results in there being negligible background noise at all quality levels.

CD-| Audio

Level Name Bits per | Frequency | Percentage of CD-I
(Requirement) sample | Response |of Channels} datastream used
CD Digital Audio 44.1 20
16B PCM kHz a2 kHz | | Stereo sig
(Super HiFi) |
CD-| ADPCM
Audio-Level

A (HiFi music mode) 37.8 8 17 2 stereo 50%
(Equivalent to LP) kHz kHz 4 mono 25%

B (HiFi music mode) 37.8 4 17 4 stereo 25%
(Equivalent to FM kHz kHz 8 mono 12.5%
broadcast)

C (Quality speech 18.9 4 8.5 8 stereo 12.5%
mode) kHz kHz 16 mono 6.25%
(Equivalent to
AM broadcast)

Chapter 5: Designing for Production 71

The critical factors in audio processing rates are the bandwidth (quality)
that is reproduced at each level and the percentage of the data stream that
is used.

A-level audio has a flat frequency response up to 17kHz (the limit of
most discerning ears), while using only 50% of the data stream in stereo
playback. C-level stereo, using only 12% of the data stream, is still able
to reproduce excellent audio quality for most television applications. It
is worth noting that B-level has the same wide frequency response as
A-level but gains a slightly rougher quality by being stored in 4 bits
instead of 8 bits.

Data Transfer Channels: Video Playback

Video data is transferred from the disc to the Image Stores before display
on the screen. Video data must be mixed into the data stream with the
audio and any software data that may be needed. The critical point to
keep in mind for video processing is the time it takes to move a single
image from the disc to RAM, given that it can only occupy that part of
the data stream that is not used up by other data.

Both DYUV and 7- or 8-bit CLUT NTSC images require 85k bytes per
full screen image. If no other data is passing through the channel at the
time (i.e. no audio), it will take 0.5 seconds to load each image. This is
calculated by dividing 85k bytes by the rate at which data can be
transferred from the disc - 170k bytes per second. But if A-Level audio
is playing, 50% of the data stream is already occupied, so it will take
twice as long - approximately one second - to load each image.

[9] opaque

[__] TRANSPARENT

Chapter 7: CD-I
Decoder: Video Processor

Chapter 3: Motion Vidco

72 CD-I: A Designer's Overview

Chapter 7: CD-I
Decoder: Random Access
Memory

Chapter 6: Hot Shot
Sports: Audio

Creating motion requires the video screen to be refreshed around 15
times per second. Partial screen updates or run-length coding techniques
reduce the amount of data required to store a single image in the motion
sequences.

Each image can occupy only 170k bytes (the data rate) divided by 15
images/second, or about 11k bytes. Since a full screen DYUV and CLUT
image requires 85k bytes in NTSC, then only about 13% of the screen
can be occupied by moving images.

However, a series of software coding techniques have been developed
that increases the amount of screen area available for partial updates from
13% to about 50% - and permits C-Level stereo sound to be played at
the same time. These techniques effectively redistribute the coding
parameters in such a way as to produce reasonable quality moving
images instead of high quality stills. Using chroma key to confine the
updates to the front plane can result in full-screen, full motion video with
synchronized stereo sound.

System RAM

Aside from the permanent storage space on the compact disc itself, there
is a temporary storage space within the CD-I system called System
Random Access Memory (RAM), which offers a total memory space of
1024k bytes or 1 MB, divided into two banks of 512k bytes each. Data
transferred to System RAM can be retrieved much more quickly than
data which must be retrieved from the disc in real-time. As storage is
temporary, everything transferred to System RAM is lost when the player
is turned off.

Operating software in the CD-I system will use some part of the available
System RAM space. While the main CD-RTOS operating system is
contained in the system ROM, about 50K is loaded into system RAM
when the player is turned on.

Application software requirements will vary according to the nature of
the program, so a simple program may only need a few kilobytes of
whereas a complex database retrieval program may use as much as
several hundred. Some of this software will be loaded into the System
RAM when the application begins, while other sections may be loaded
and unloaded as the application progresses.

Most music and voice-over audio will be retrieved from the disc and sent
directly to the Audio Processing Unit (APU), without using any RAM.
However, some types of sound may be stored more accessibly as
soundmaps in RAM for use at cued moments in the program or to await

Chapter 5: Designing for Production 73

the user’s actions: for example, a sound effect which signals a correct
decision ina game. Effective use of soundmaps can help to mask the time
taken for a disc seek.

Memory requirement for a soundmap is calculated on a percentage of Chapter 3: Audio:
the data rate. The memory needed for a sound effect which lasts for 3 Soundmaps
seconds at C-Level stereo is calculated by taking 12% (C-Level Stereo)

of the data rate (170k bytes x 12% = 20K bytes) and multiplying by three

seconds - about 60k bytes altogether. Once this is transferred to RAM,

it can be used as often as necessary in the course of the program without

recourse to the disc.

Video is the most critical user of RAM storage data as all visual images Chapter 7: CD-I
must be loaded into RAM before being displayed on the screen. Each Decoder: Random Access
bank of 512k bytes supports one video screen plane. Memory

Effects like dissolves or wipes require the second of the two pictures in
the effect to be held in RAM and brought into play when needed. DYUV
images use 90k (105k PAL) or about 20% of the available RAM space.
When these images are loaded into RAM, their display timing is
synchronized to the flow of the audio track, then they are unloaded to
make room for new images.

Judicious design thinking is required to make efficient use of the limited
RAM space available for both sound and pictures.

Main Processor Power

Video images can be synthesized or manipulated in the MPU before Chapter 6: Hot Shot
passing to the screen planes for display. Drawmapsare spaces inmemory Sports: Action Areas
that can be manipulated by the MPU. In the golf game described earlier,

the direction and size of the ball travelling down the fairway is

determined by the conditions of swing and the timing of the hit that a

player signals to the computer. A drawmap of the golfball held in RAM

is manipulated by the MPU and a trajectory is calculated with data

received from the user interface. Progressive variations in ball size and

screen position are calculated and displayed as the ball appears to move

down the fairway from the tee, rising in flight to meet the viewer’s eye,

then descending once again to land by the green in an excellent position

for the second shot.

The CD-I designer is faced with the task of creating applications for a
consumer market whose expectations are based on broadcast television
standards for sound and picture. The only real limitation to
accomplishing that is the confined data flow, which hampers the use of
unlimited full-screen full motion video. Managing the complicated data

74 CD-I: A Designer's Overview

storage and processing power restrictions within CD-I will be dealt with
in the next section.

THE MECHANICS OF CD-I DESIGN

One key to effective design for CD-I is controlling the display of
information in sound and pictures by managing the storage and transfer
of digital data in the CD-I system.

CD-I is an interactive medium: the design of audio and visual images
must be directed to the activities of the person using the application.
Interactivity is not yet common in home entertainment and home
education, so designs must allow for the consumer’s initial unfamiliarity
- and, perhaps, nervousness - with the concept. People may not want to
be constantly making choices, answering questions, looking for
information or even playing games: they may want to watch a bit of
television now and again.

These sorts of consideration should be allowed for in program design.
The opportunity for a user to enjoy a variety of program styles will make
successful programs. This section explains the various factors that must
be brought together to present program material to the user.

Interface Devices

While the choice of interface devices of course varies widely across the
range of potential applications and the degree of interactivity in any one,
a basic device will have two buttons and a pointer to control a cursor on
the screen. This might typically be a remote control pad, though other
simple pointing devices might also be used. More complex devices might
include a computer keyboard, a music synthesizer, a graphics tablet, and
even special tools like light pens or bar-code readers.

Screen Design

js “S97 as

# | pfs S

Chapter 5: Designing for Production 75

It is inevitable that expectations for this medium will be based on
broadcast television standards, not only for picture quality, but also for
pacing and many visual and conceptual elements. CD-I applications
hardly need be modelled after trend-setting TV shows, but some
awareness of the conventions of that pervasive medium is important.
CD-I does offer many screen effects that help to create the visual style
of broadcast television.

Magazines and television advertisements are also good models for CD-I
screen design: the screen can be regarded as an electronic magazine page
on which elements of text, photos, graphics and even full motion video
can be combined for a variety of effects. And while one of the limitations
of CD-I is that full screens of information require large amounts of data,
it is not always necessary to change the whole screen at one time: small
areas of change will accelerate the pace of the presentation without
straining the data flow capacity. Ultimately, pacing depends on the
frequency rather than the volume of change.

Screen Effects

Screen effects - what the user sees - are the visual palette of the CD-]
designer, and require management of the different types of image coding
to be displayed on one or both of the two 8-bit image planes.

Image coding is explained in Chapter 3. What is important here is the
relationship between images in dual plane configurations and the
restrictions that govern which planes are used to display certain types of
image.

Dual-plane DYUV

DYUV images can be shown on both Plane A and Plane B, and these
planes can be reversed while mixing images. However, problems of data
rate arise in fast scrolling, and partial update effects (explained in Chapter
3) are difficult to achieve with natural images in CD-I. The subtlety of
color coding in DYUV (and 8-bit CLUT) make it difficult for the system
to calculate the necessary starting values, or their locations, for rapid
updates to happen properly.

Chapter 3: Visual Effects

Chapter 3: Visual
Effects: Scrolling

76 CD-I: A Designer's Overview

CLUT over DYUV

__}] TRANSPARENT

In this configuration, the 8-bit 256-color CLUT is taken up entirely in
the foreground or Plane A, leaving Plane B to be encoded in DYUV. As
the Color Look-Up Table is completely used by the foreground image,
no space is left for any other CLUT-coded graphics. As RGB images
require a full 16 bits, a single DYUV image is the only possible image
that could be displayed on Plane B. The CLUT/DYUV combination
creates a high quality natural background, such as a landscape or interior
room, with a flexible CLUT graphic in the foreground. In a language
‘program, a graphic plate with a customer’s bill could be shown over a
DYUV image of a French cafe scene to show through. Plane A could be
partially updated to show changes in the bill if the customer wished to
dispute the total.

The only advantage in using 7-bit CLUT in the foreground would be to
increase the animation speed of the graphic by Run-length coding.

This configuration would be used to present text information on Plane A
in front of a DYUV image. Text or simple graphic icons in a children’s
entertainment program would benefit from double resolution without
being affected by the reduction of color choice. 4-bit CLUT would also
be useful for simple high-speed animation. -

Dual 7-bit CLUT

This is a very flexible image configuration. CLUT graphics can be
displayed on either screen. Scrolling is less limited with dual 7-bit CLUT,
and restricted only by available memory space.

Chapter 5: Designing for Production 77

3- or 4-bit over 7-bit CLUT

This configuration is best suited to foreground text or simple foreground
animations over a more subtle graphic background. Both 3- and 4-bit
CLUTs have a narrow color range and are best suited to large areas of
animation. Remember that 8-bit backgrounds cannot be used in
combination with other CLUT planes.

It is possible to use a 4-bit plane in the background if simplified colors
are appropriate in both foreground and background - for example, for a
text menu over a simple colored or fast-scrolling background page.

Dual DYUV with CLUT subscreen

In this combination, part of the screen appears as ordinary DYUV
images, but another part is designated a sub-screen running the full width
of the screen but only part of its height.

This sub-screen can be used for a complex graphic panel to control of | Chapter 6: Country
the changing images on the remainder of the screen. For example, by House Murders
using parts of the graphic toolkit, the detective in a mystery game could

control the path of a surrogate ’ walk’ through a country house where the

murder has been committed, interview suspects or gather clues.

Partial screens, requiring less data storage, save memory space in the
graphic area and reduce the size of natural images to be updated in the
DYUV part of the screen.

Controlling the Dataflow

The fundamental challenge for CD-I design is to bring together the
design skills of existing media such as interactive video, television or
book design, with the management problems of digital storage space and
data flow particular to CD-I. To visualize the scope of possible
applications, the CD-I designer needs to be at home with the calculations

78 CD-I: A Designer's Overview

Chapter 7: CD-RTOS

which keep track of CD-I data. The design factors discussed here are
applied to potential applications in the next chapter.

Interleaving

Information stored at the beginning of each sector tells the CD -
Real-Time Operating System what that sector contains - the type of data
(audio, video, text, software), whether that data is real-time, to what
channel the sector belongs, and so forth.

CD-RTOS directs each sector in turn to its appropriate processing
location within the player. It acts like a traffic controller, directing each
vehicle in the linear stream to the correct lane in order to make the most
efficient use of space available. However, the traffic is not random since
the CD-I program designer has some control over the types and sequence
of sectors in the CD-I traffic flow, and thus the efficiency of CD-RTOS.

A-LEVEL AUDIO

In the diagram above, A-Level stereo audio is to be played with a series
of dissolving photographs of a neighborhood in Paris. Audio of this
quality requires 50% of the data stream, but for the playback of the audio
to be timed correctly, each audio sector must be followed by a single

Chapter 5: Designing for Production 79

sector of other data such as the video data or possibly text or application
software.

In this case, it is the audio that determines the speed of picture change,
and the pictures are synchronized with the audio channel. Video requires
approximately 37 sectors (SOPAL) to load each DYUV image into RAM.
Of the 75 sectors per second picked up by the player, half are already
occupied with the audio - leaving only half (or 37 sectors/sec) for video.
At that rate, pictures are available at intervals of 1 second. In an actual
photo essay, it is reasonable to assume that pictures do not have to change
more than once every three to five seconds. Furthermore, the full screen
need not change every time. Thus, this combination of audio and video
data is well within the capabilities of CD-I.

If this application were part of a large database like an encyclopedia, or Chapter 6: Grolier
a language phrasebook, then total storage space on the disc could affect Multi-media

the calculation: for example, C-Level stereo, rather than A-Level, would eed clopedia
provide adequate sound quality and leave more disc space for other types — french Phrasebook

of information.

C-LEVEL AUDIO

80 CD-I: A Designer's Overview

Chapter 6: Pop
Showcase: Disguising
Disc Seeks

If C-Level stereo is used, only one out of every eight sectors in a sequence
is required for the soundtrack. If pictures change by dissolving or cutting
every four to six seconds, then other sectors in the data stream can be
allocated to additional information that may be requested by the user,
including completely different C-level mono sound.

Seek Time

The value of keeping a variety of programme choices grouped together
in the data stream is the speed with which the CD-I player can respond
to a user’s request for change. If the data is not present in the data stream,
the player must seek it in another part of the disc, which is likely to disrupt
the programme.

Seek time is the time needed for the pickup head of the CD-I player to
move from one part of the disc to another, for the motor to adjust its speed
accordingly, and for new data to be read and decoded in the system. CD-I
discs revolve at variable rates depending upon the part of the disc that is
being read - that is, they have a constant linear velocity (CLV), or
constant data rate of 75 sectors per second. Moving towards the outside
of the disc, the motor slows down; moving towards the inside, the motor
speeds up. Picking up data stored within a 20 Megabyte (Mb) range (or
within 3% of the disc) does not require the motor to change speeds
significantly, so seek time is negligible. The total time to access new data
may still be as much as one rotation of the disc (as much as 1/4 sec). At
170k bytes/sec, 20 Mb amounts to almost 2 minutes of real-time play or
between 100 and 200 full screen DYUV images depending upon the
quality of sound interleaved with the images.

Unlike interactive LaserVision, sound and picture can be stored in RAM
and played through the system at the same time that the laser head is
moving to another location on the disc to find the next sequence of
material. Most seeks can be disguised in this way.

Synchronization

Various segments of CD-I program material travel along different paths
and undergo different processes. Each type of data (video, audio, text,
software) is broken down into units of about 2k and stored in sectors
along the disc track. Thus, the playback of these sectors to the video and
audio receivers must be synchronized to ensure that the audio track lines
up with the right picture and that special effects in video or audio are
correctly timed.

Within the data of any sector is a trigger bit, which tells the application
software that this is a synchronization point and that something specific
such as a special effect should happen here.

Chapter 5: Designing for Production 81

Synchronizing to Audio

In synchronizing with audio commentary, the video picture may change
on a word or between specific phrases. The trigger tells the application
software the exact time that the specific sector is being read by the player
so the software can then cue the image change or any other effects that
may be stored in RAM or generated in the MPU for that moment.

Synchronizing to Video

Ball Strike

C-level mono Ball Strike Club Swing —_ Audio
Golf Course Ambience. Audio to RAM. Audio Playback
par an —_—_—,_—,

Strike Fairway

Fairway Golfer Foreswing
Trigger Scroll.

Graphic Character Animation Animation

The task is similar to that already practised in motion film or video. At
the start of a sequence the audio must be correctly aligned with the video
to ensure that a speech, for example, synchronizes with the lip
movements of the speaker. Once the correct alignment has been made,
sound and vision will proceed at a constant rate.

In CD-I, when the user may intervene to determine when a particular Chapter 6: Hot Shot
action takes place, the task becomes more complex. In a game like the Sports: Audio

golf simulation introduced earlier, the visual of a golf club swinging must

synchronize with a carefully-timed audio effect as the club hits the ball.

The timing of the synchronization cue depends on the instant that the

player - acting through a remote control or other input device - designates

as the exact moment the club head contacts the ball. This type of effect

can be handled economically as a soundmap pre-stored in RAM and

played through the audio output channels at the required moment.

Timed Cues
The CD-I system has an internal clock which can be used to generate Chapter 7: CD-I
timed cues such as, for example, a limit on user response-time: if nothing Decoder: Clock/Calendar

82 CD-I: A Designer's Overview

happens within the designated period, the software proceeds to the next
action on its own, perhaps advancing the programme, or perhaps shutting
the system off.

Interactive Design

Interactivity covers a broad range of possibilities depending on the
degree of user participation required. At one level, interactivity could
keep the user guiding the program and responding to cues. The attention
is active and highly engaged. Depending, of course, on the nature of the
content, a linear sequence in an interactive application is usually best
kept shorter than 20 or 30 seconds, before some kind of user action is
required.

CONTROL
GRAPHIC

KNOWLEDGE
EXPLORER NCYCLOPEDIA
ESSAY

At another level, interactivity may require the user to select a picture
essay on Science, Art or Geography from an encyclopedia. The essay
might then play for 5 minutes unless interrupted by the user to branch
into other essays, or into a more interactive search through text
information.

CONCLUSION

This chapter has covered the basic technical principles of CD-I design.
Along with Chapter 4, a comprehensive view has been given of the
Design Process from the initiation of the Brief, through the Idea Map, to
the development of various levels of Storyboard. Chapter 5 looked into
particular technical constraints that are specific to design problems in
CD-I. Using this as a groundwork, the next chapter will present several
detailed examples of CD-I applications that illustrate various design and
data management concepts.

Chapter 6: Typical CD-I Applications 83

CHAPTER 6: TYPICAL CD-I APPLICATIONS

The previous chapters outlined the nature of CD-I multi-media and the
factors which contribute to good application design. This chapter looks
at some specific design concepts and how individual applications
discussed in other chapters illustrate these concepts.

It is difficult to find subjects which do not lend themselves to
development as CD-I discs. What may be even more difficult to realize
is that when existing products are converted to CD-I they become entirely
new concepts, quite different from the printed work, audio-visual
program or computer software on which they were based.

An encyclopedia, for example, is no longer a series of articles, long and
short, in alphabetical order, with pictures to illustrate a few. It may begin
with a choice frame offering several voyages of exploration through
picture, sound and text databases. You can study any topic you choose
in a variety of ways - gaining a general overview in a short audio-visual
essay, perhaps, before going on to see what more specific information is
available, and choosing an area to study. You can pick a word - your own
surname, for instance - and see how many times and where it is
mentioned in the text database. You can even select information such as
a picture, text and audio commentary, to create a short presentation of
your own.

Similarly, the experience of watching a classic film, play, dance work or
opera, for example, can be greatly enhanced through CD-I. The disc
might contain, as well as the performance itself, critical essays about the
work, short notes on key passages (which you can call to the screen at
the appropriate time), interviews with key personalities, biographies of
the principal performers and production team, and background
information about the making of this production. You may listen to the
original script or libretto, and call up sub-titles in the language of your
choice. You can even dub out the stars and take over yourself, and make
your screen debut opposite Garbo, Olivier or Callas!

The following are just some of the topics that lend themselves to CD-I.
They certainly show the range and diversity of the CD-I publishing
opportunity.

® pop music, movies, plays, dance and opera, enhanced as described
above

* studies of famous people and events in history and popular culture

84 CD-I: A Designer's Overview

art and music programs which allow the user creative control
games of observation/deduction, such as mysteries and adventures

educational games for children, to teach learning and social skills as
well as academic subjects and knowledge areas

interactive movies and even erotica which allow the user or player to
direct the action

games of skill such as bridge or chess, or enhanced versions of board
games such as Monopoly

multi-media reference works such as encyclopedia and dictionaries

diagnostic reference books on specialist topics from family medicine
to car repair

picture libraries and databases for amateur and professional collectors,
scholars and hobbyists

games of general knowledge, wit and experience, such as trivia and
word games

armchair travel guides and tourist books

guides to famous places and buildings, from archaeological sites to
museums

maps, plans and navigation aids - including in-car systems
*surrogate travel’ through fabulous places (real or imaginary)

arcade-style games demanding hand/eye co-ordination and quick jud-
gement

educational material at all levels from pre-school to post-graduate
language teaching for self-tuition or institutional use

industrial and commercial training material, both off-the-shelf and
made-to-order

catalogs and sales aids, for use by customers and in co-ordinated sa-
les presentations, and for staff training

And these are only a few ... A videotape has been developed which
simulates some of these applications, keeping entirely within the ’Green
Book’ specification. Chapters 4 and 5.developed a notion of the Design
Process through three main stages - The Brief, The Ideamap, and The
Storyboard. The following example applications illustrate various
aspects of CD-I design issues presented earlier. The encyclopedia deals
with the major issue of total storage capacity for one project on a single

Chapter 6: Typical CD-I Applications 85

disc. The Hot Shot Golf game will show an example of an early stage
storyboard, where ideas are sketched out for such design considerations
as screen style, program sequencing and possible interactive branching
routes. The French phrasebook will show an example of a later, more
specified stage in the storyboard process when the overall design is
complete and ready to be released to production.

THE GROLIER MULTI-MEDIA ENCYCLOPEDIA

The world’s largest publisher of encyclopedias, Grolier, has already
identified CD-I as the logical choice for a multi-media edition of their
Academic American Encyclopedia. Grolier is currently designing the
world’s first interactive encyclopedia on CD-I because CD-I provides an
immense capacity for the storage of text, pictures and sound and because
CD-Iallows the user to browse interactively. Even when stretched to the
maximum search time, the CD-I encyclopedia will call up information
much more quickly (and of course more accurately) than a reader can
with the 20-volume original print version. Furthermore, CD-I can
illustrate the text database with sounds and pictures, audio-visual
experiences that add an entirely new dimension to information retrieval.
The Multi-media Encyclopedia will tempt passive viewers to explore
deeper layers of information through interactivity.

Designing the Interactivity

The key question in designing an encyclopedia is to be very clear about
how the interactivity will work. Rapid and accurate text retrieval is an
early priority, but many people love just to browse, and to have their
curiosity stimulated by what they come across. Yet there must also be
opportunities to take a break from the interactivity and be entertained.
Audio-visual essays have been designed to provide hours of
entertainment by taking the viewer on journeys around the vast amounts
of knowledge by combining the text, pictures and sounds into short
presentations.

Mode of Use

The text, picture and sound databases provide a natural foundation for
the development of an overall structure for the various types of
information to be contained on the disc. This will lead naturally into
development of the first level of interactive branching to retrieve
information from each of the areas.

With the Multi-media Encyclopedia, information will be divided into
several inter-connecting databases (’Domains of Knowledge’) in a web

86 CD-I: A Designer's Overview

that enables the user to move freely among them. These databases
include:

° A series of audio-visual essays on general topics within main catego-
ries such as, for example, Arts, History, Geography, Science, Sports
and Ideas;

* An audio database of speeches, sound effects, music and song;

° The full encyclopedia text itself - over 10,000,000 words fully index-
ed.

Data Management

The critical data management issues for an encyclopedia concern
absolute disc capacity rather than dataflow rates. However powerful
CD-I may be, and however large its storage capacity, the megabytes still
get used up awfully quickly when you are trying to cram an overview of
the whole of human experience onto a 12cm disc!

Screen Interface

To allow entry to these various Domains, two types of screen interface
are essential. One enables direct access to any of the Domains from the
moment the disc is loaded. The other allows access from within one
Domain to any of the others - a lateral structure. Each must be compatible
with a simple input device like a mouse or tracker ball.

Menus

Menus provide instant access to any of the Domains so, for example, the
user can pursue a specific topic or question. If the user simply wants to
browse, the Knowledge Explorer offers a choice of short introductory
essays on general topics. The menu here cycles a series of still images
from each of the essays in the appropriate box, which actas visual stimuli
to suggest what each essay contains.

Chapter 6: Typical CD-I Applications 87

Choosing just one of the essays begins a sequence of passive television
viewing that can continue as long as the user wishes. The interactive
aspect is up to the user to choose, otherwise the encyclopedia offers linear
video without requiring constant or even intermittent attention.

Screen Effects

The short essays use the screen effects within the CD-I player to show
photographic DYUV natural images and CLUT graphics.

Dissolves, wipes and screen montages give a television style to the
presentation; however, as these effects are generated within the CD-I
player itself, the individual images remain untouched, and can appear in
a variety of forms - in the essays, or within any number of picture bank
on various subjects.

This application could contain several thousand separate images, each
of which could be used in several ways.

Graphic Control Panel

After simple access menus, the second type of screen interface is a
flexible control panel which allows various degrees of interactivity.

If, for example, the user stops one of the Knowledge Explorer’ essays
at any point, the image on the screen freezes and the ’control panel’
appears as a graphic and matte over the picture.

The control panel is coded as a 7-bit CLUT displayed on Plane A. Since
the panel can be called up at random, the software must ensure that the
image from the audio-visual essay is moved to the background if it is not
already there.

Interactive Branches
Several areas of the control panel are active regions’ and correspond to
directions that the user may explore.

Touching the option ’Caption’ brings up a caption in the transparent area
of the foreground panel. This contains additional information about the
picture or subject on the screen. Each caption is interleaved with the data
for its associated picture.

In the same way, the ’Links’ button brings up a list of options for
connecting to other parts of the encyclopedia. This list is also interleaved
with its associated picture.

88 CD-I: A Designer's Overview

?UTAMKHAMEN

AINE OF ELVPT

Although the pictures in the essay change slowly, and use C-level stereo
for background audio and voice-over, the data stream is also occupied
with other data elements which may not be used every time but which
must be available to the user.

Domain Seeks

To reach some of the Domains offered by the Link option, or to research
in the text Domain, the player pickup would have to move to another area
of the disc, using a delayed seek time. This is acceptable as a major
change is being made.

The CD-I Multi-media Encyclopedia is able to hold a vast amount of
information on one disc: up to 10 million words, 3000 pictures, and three
hours of sound - enough for a complete 20 volume printed encyclopedia
with audio-visual essays and interactivity as well. However, the large
volume of data being stored, and the complexity of access, requires
careful management and design.

HOT SHOT SPORTS

The treatment for Hot Shot Golf has already been described in Chapters
2 and 3. Athletics is another topic that lends itself well to inclusion in the
CD-I range of applications. The Olympic Games give pleasure to
millions of television viewers around the world, but imagine how their
pleasure can be enhanced even further with the CD-I Athletics database.

Chapter 6: Typical CD-I Applications 89

In the early stages of the High Jump competition, you can park the
television picture carrying a live transmission from Seoul and call up the
High Jump database. Who is the record holder ? What happened last time
there was a jump off for gold in the Olympics ? Study the form of the
contestants. Use the Slo Mo’ feature to step frame by frame though
motion pictures of the last finals. Call up ’Chalk Talk’ and have the CD-I
coach talk you through technique. So when the time comes for the jump
off, your pleasure is greatly increased by the insight and understanding
you have gained from Hot Shot High Jumps.

Back to golf. This application makes use of interactive computer
animation as well as high quality photographs and ambient background
sound. What is significant about this version of the game is that the
graphics of the computer arcade game are combined with photos of areal
course to allow the viewer to play alongside the professional. In fact, the
storage capacity of the disc is sufficient to offer the player a choice of
famous courses as well, from Augusta to St Andrews. The animation
program for the game would use the same factors no matter which course
was selected.

The accompanying outline storyboard shows the earliest stages of idea
mapping, with sketches of a key image for each sequence of the game.
These indicate the main types of plane configurations, sound quality, and
the transition type (that is, whether the scene is part of a linear sequence
or a pivotal frame allowing branching to other parts of the application).

The production tasks then fall into two areas. The first is to accumulate
a database of photographs covering a representative sample of possible
positions from which shots could be taken as a player moves through
eighteen holes. The second is to develop an interactive game animation.

Photographic Database

Each game needs about 750 photographs, shot specifically for this
project. For each of the eighteen holes, there is a tee shot plus five
separate views spanning the fairway at 100 yard intervals, and a good

90 CD-I: A Designer's Overview

cprprosncrtitin: Hot Shot Sport sheer fg

Sequence: Main Chose Frame
Plane A: CLuT-7 - Matte
Plane B: Cuur-7 - tat serell

Source: Commission Graphic
Time on screen: User Choice - max-hold 66 Sec
Audio: b/map-theme 5 ste /sff 60sec

C- Stereo
FX:

Transition: Branch - Sport menu - lursr On

Sequence: Match play Choyce frame
Plane A: CLuT-8 - matte
Plane B: Druv - still - 25%.

ice)

Time on screen: User thowce - max 305%
Audio: ok

FX:
Transition: Branch - Course Intro

Sequence: (surse Module 1: Augusta
PlaneA: = CLUT-@ = - 100%, ae
PlaneB: Off

cig Some Gommission Crephe
Time on screen: User Choice - max 66sec.
Audio: C- Moho

FX:
‘ : ~ Hole Caddy
Transition: Branch Be nea se

- Return fo Choices
Sequence: Hole Module - /S

Plane A: CLur -7 - menu 20%
Plane B: dDYuv -(SStee - (00%

FE Source: Commisscan Graphic/Libary Photo

H Time on screen: User Chotee - max hold 30

Audio: C- mono ambience

FX: birds

Transition: - Game
Branch > Pro shop

ea eee 5 (+

Chapter 6: Typical CD-| Applications 91

corproucrirtie: Hot Shot 5 port sheer 1D
ie Soquence: Pro Shop - (5% hole

TE PlaneA: dyuv -/S* tee - /00%o

PlaneB: DyuV- sandtrap - 50% hey

Source: Commission Photo

Time on scresn: 4 see (mage

— C-skree
V0. - Commentary
FX: Ambience

Transition: /jpear disselve/return fy Hole Mod.

1 Sequence: Course Histo
PlanoA: SyYUY Nick Lbas -756%125o
PlaneB: § cLuT-7 Facthie fest

Source
Time on screen: £ Sec image
Audio: C- stereo
V.0.t¢ muste
PX: crowd cheer

Transtion: 7 ineay disselve [return ble mod -

Sequence: Game -Ana:mation

PlaneA: CLUT-3 Golfer

PlaneB: § ceut-7 -/5%Hole/vertical scroll
groph«c

SB Source: HPu sottwove | Commission avaphic

Time on screen: Lateractive . max hold 26sec
Audio: C- mono ambience

FX: suing [hit [crowd

Brand. return / Itnear ty neat
osttisn - bYuv

Transition:

Golfer Animation

Start Back swing Stayt Foreswing Ball Contact

92 CD-I: A Designer's Overview

selection of photographs on and around the green. Each photograph
should give the player a good view of the next shot from the point at
which the last shot ended.

These photographs can also be used in an informative way: the caddy
menu, for instance, besides dispensing the chosen club, also offers advice
from the club professional on the best way to play that particular hole.
Short documentary sequences could relate anecdotes about famous
golfers or famous tournaments at the course.

Screen layouts consist mainly of DYUV images of the course on the
background planes with graphics of menus or the animated figure of a
golfer on the foreground.

The Animated Figure

The golfer is an interactive animated figure which is handled by the
microprocessor within the CD-I player (and remains the same no matter
what course is represented in the photographs). The figure is seen from
above (unlike the figure in the popular computer golf game) to allow the
computer to spend its power on interaction with the user rather than ona
complex graphic. With Hot Shot Golf, the player actually determines the
flight of the ball by controlling the swing.

The golfer is a 3-bit CLUT animation in the foreground. In the
background, the DYUV image looking down the fairway is changed to
a 7-bit CLUT and replaced in the centre with a 7-bit graphic depicting
an aerial view of the hole which will scroll vertically when the ball is
struck.

Changing the DYUV view to CLUT keeps the background plane in the
same coding system and CLUT scroll alleviates some of the difficulty of
DYUV scrolling.

Chapter 6: Typical CD-I Applications 93

Action Areas

Two ’action areas’ have been determined for the golfer. The first is the
hat: placing the cursor on the hat and holding the button down on the
remote control device allows the player to align the golfer’s feet and thus
the direction of the ball, taking into account factors of wind direction.

The second action area is the ball: when the cursor is moved to the ball
and the button held down, the golfer’s backswing is activated; when the
button is released the foreswing begins; at the moment the player sees
contact between the lub and the ball, the thumb button is tapped once
again.

Ball Strike
C-level mono Ball Strike Club Swing Audio
Golf Course Ambience. Audio to RAM. Audio Playback
aS po yt

: Ball
Fairway Golfer Backswing Foreswing Strike Fairway
Graphic Character Animation Animation Trigger Scroll.

94 CD-I: A Designer's Overview

This is a point of critical timing, both for the player and the designer. The
microprocessor has to deal with several calculation factors. One is
direction, based on alignment of the golfer’s feet and by the point of
impact: touching the button too early causes it to slice to the right,
touching too late hooks it to the left, and correct timing sends it down the
middle.

Another calculation is speed and distance of flight, which depends upon
the amount of backswing and the club used. The microprocessor
calculates these factors and displays a ball, which has been stored as a
drawmap in RAM, increasing and decreasing in size as it flies down the
fairway, coming to rest before the next shot. The speed of the scrolling
fairway is co-ordinated with the speed of flight, so a poor shot may fly
off the fairway to one side.

Audio

As audio is not critical for quality, C-level mono is used for a background
of birdsong and the like. However, sounds such as the swing of the club
through the air and the impact on the ball make the moment of contact
more vivid. Four sound effects are loaded as soundmaps at the beginning
of the swing: the loop of the foreswing cutting the air, the sound of the
club hitting the ball, and a choice of crowd sounds appropriate to the
quality of the shot - either a round of applause or a murmur of
disappointment!

Of course, this type of interactive animation against a realistic
background could be developed for virtually any single-player sport and,
with some imagination, for team sports as well, with the user taking the
role of a key player or even the coach.

COUNTRY HOUSE MURDERS

The murder mystery is a classic game, well suited to CD-I. The
combination of interactivity and multi-media allows the designer no end
of opportunities to thrill and suprise the audience. In the version designed
for the demonstration tape, one or more players act as detectives trying
to solve a murder in a stately country house; the computer randomly
selects details of the crime, which change with every playing.

Surrogate Travel

The atmosphere of the house is created through a technique known as
*surrogate travel’. The house has ten main rooms plus corridors,
cupboards and staircases. As the camera travels through the house, a
series of photographs is taken, at eye level, and at carefully graduated

Chapter 6: Typical CD-I Applications 95

intervals, to simulate the impression a real visitor would have wandering
through the house and in and out of its many nooks and crannies. These
DYUV images will ’cut’ in sequence on the background plane as the
detective moves about the house. There is no jumping from room to room
here: many things can happen to unwary detectives in the poorly lit
corridors of old houses!

Movement is controlled by a CLUT direction graphic on the foreground
plane. Touching the arrows allows the player to turn left or right or to
walk forward.

Handling Disc Space

Storing photographs on the disc is a problem of disc geography. Images
are sequenced so that the cuts can happen quickly and smoothly. Turns
may require a short seek to the start of a new series of images. It would
be preferable to jump at this point rather than after the turn and the start
of the new ’forward’ sequence.

Possible false turns, where no further forward motion is possible, could
be interleaved into the series: for example, the detective could stop in a
corridor to look at a painting which might contain clues... or fall down
the cellar staircase by taking too sudden a turn.

Icons

Routes through the house and the task of clue gathering is helped by
various graphic icons in the detective’s ’tool kit’. At the start of each
game, the detective is offered a choice of potentially-useful objects to

96 CD-I: A Designer's Overview

put into the kit - of course, there is not room for them all. For example,
a plan of the house (which can be overlaid on any frame to show the
player’s present location) would be useful- and so would a flashlight.

Other tools might include a cassette tape-recorder for interviewing
witnesses and suspects, a fingerprint kit, a notebook and even clues
gathered during the course of the investigation. As with many adventure
games, the detective may not be able to carry everything at once, and
may have to decide what to foresake, and where to leave it.

Audio

Audio quality is not critical in this game, but could be used for interviews,
or random special effects like squeaking hinges and floorboards, or
distant cries of distress, to add flavour to the hunt. The game might offer
different levels of interactivity with, say, special sound effects to draw
attention to important clues or even hints for the youngest or slowest
players!

POP SHOWCASE

This application is about pop music, so the sound quality is a critical
design issue. Pop Showcase is designed in two parts: the first
concentrates on the music of a particular band or performer, the choice
of A-Level stereo for optimum listening quality; the second is an
information bank with details of the stars’ background, greatest hits,
tours, gossip and so forth.

Chapter 6: Typical CD-I Applications 97

Since A-Level sound uses fully 50% of the data stream, the design of the
accompanying screen displays is critical, particularly as the screen must
seem to move frequently in keeping with the general tempo of the music.

Partial Screen Updates

A magazine style was chosen both to suit the lively tone of the program
and to meet the practical need to keep the update areas small. As a
particular song plays, DYUV images change in different areas of the
screen as a pictorial collage to accompany the music, and also as a menu
or table of contents for other material contained on the disc.

Disguising Disc Seeks

Highlighting any one of the images on the menu will stop the song and
lead the listener into various related information areas, which might
include a biography of one of the band members, or a scrolling list of hit
records.

The music is used as background sound so the audio quality can be
reduced to C-Level here as the graphic and DYUV components of the
screen demand a larger portion of the data stream - and the viewer’s
attention.

Moving from the Menu to specific information requires disc seeks. To
keep the audio playback alive, the pickup head plays bursts of sound as
it skips to a new area of the track. This is a way of disguising seek time
- by making it audible!

98 CD-I: A Designer's Overview

Singalong
A feature of Pop Showcase is a singalong section: as a song plays, the
lyrics are displayed line by line with a bouncing ball that keeps time to
the music.

The magazine format of the screen is maintained here, and greater
interest is added by incorporating segments of CD videos coded in CD-I
digital form onto the background plane. As the listener is probably
singing, and not listening too closely, audio quality is again reduced to
C-Level stereo,which leaves maximum room for the DYUV full-motion
video and CLUT graphics in the stream.

However, as full-motion video is intended here, some careful
calculations must be made to determine how the effect is presented to
the viewer and what proportion of the screen can be occupied.

Screen Proportion

The foreground plane is a 7-bit CLUT graphic. It holds the scrolling
texture plane with a matte through which a mixed DYUV and CLUT
background in the other plane is visible.

The second plane through the matte is then divided into two subscreens:
an upper partin DYUV containing part of the still pictures and the motion
video, and the lower part in CLUT together with the rest of the still, the
bouncing ball and the text. As the part of the still that is in the DYUV
subscreen was CLUT coded originally, the joint between the two parts
of the image is not visible.

The bouncing ball is created from a computer program within the CLUT
subscreen.

The computer program for the bouncing ball and the scrolling of the front
plane are pre-loaded into the program space in System RAM. The text
bars are pre-loaded as drawmaps. While running the sequence, the partial
update motion video and the sound are read directly from the disc, and
synchronized to the program.

The partial update used here is 70% of the data stream. Thus 70% of 170k
bytes/sec. yields 119k bytes per sec., which at 15 frames a sec. represents
8k (9.5k PAL) bytes/screen area/frame. In terms of screen area this is 8k
(9.5k PAL) divided by 90k bytes (105k PAL), or about 9% of the full
screen. This would be a rectangle just under 1/3rd by 1/3rd of the total
screen.

Chapter 6: Typical CD-I Applications

The sound, C level stereo, takes 12% of the data stream. This is slightly
under the full data rate, as the processor is heavily occupied with ball
animation and the scrolling effects in this example.

As the video is not critical to the singalong sequence, this size of image
would be very attractive in collage page format.

INTERACTIVE UNDER FIVES

Pattern, word and number recognition are learning concepts that all
children under five years of age work hard to acquire. CD-I offers a
unique and potentially highly effective means of providing attractive
learning games.

One-to-one relationships with parents and teachers remain important
features of learning at this age, but CD-I can provide hours of exciting
supplementary practice. Bright graphic animations using 3-bit CLUT
coding would enliven the acquisition of reading and counting skills no
end.

Shape Recognition

In this sequence, a yellow bird is sitting in a tree created as a 77-bit CLUT
in the background. A 3-bit CLUT sub-screen stretches across the bottom
holding three silhouettes of familiar animals, one of which is the same
as the bird sitting in the tree. An oversized cursor moves across a
sub-screen at the bottom of the other plane. The object is to match the
images.

100 CD-I: A Designer's Overview

The simplified cursor has been designed as very young children have
difficulty co-ordinating a pointer in the usual way: thus, the cursor is
incorporated into the graphics as a drawmap icon which moves about the
screen in fixed or random patterns. When the icon is in the appropriate
area of the screen, the child simply hits the specially designed large green
control button.

The cursor shape is loaded into RAM. The microprocessor calculates the
speed of the moving image and its position and displays the drawmap in
that position on the screen, updating the image as it moves across. Certain
areas of the screen can be activated so that, if the cursor is in that area
and the button is touched, the system is cued to carry out the next action.

Sound is not critical in this sequence, so a background tune in C-Level
mono is adequate, but for added interest, three sound effects are loaded
into soundmaps in RAM as the scene begins - one of them is happy bird
song. A voice-over acts as a guide to each game.

Each icon is active when the cursor appears in its silhouette area. If the
button is touched while the cursor is in the bird silhouette, the bird will
begin to flap its wings and sing its cheerful song.

As audio can be kept to C-Level, and the images are mainly animated
graphics, CD-I storage and data rates are not strained - and nor is the
production budget.

FRENCH PHRASEBOOK

CD-I offers a very flexible system for language tuition in the relaxed
atmosphere of home or in an educational context such as a school. It has
the capacity for simultaneous multiple language versions, live action
dialogue scenes, and of course the interactivity that allows the student to
practice in real-time simulations.

The CD-I French Lesson is patterned on standard home teaching systems
which use books and audio cassettes to combine the pleasure of learning
about a foreign culture with useful phrases for an upcoming business trip
or holiday.

This particular phrasebook contains:

® typical conversations in cafes, streets, stations and the like;

® special phases for familiar situations such as asking for directions or
changing money;

Chapter 6: Typical CD-I Applications 101

° the opportunity to learn new words simply by touching specific ob-
jects in a scene;

* accumulation of useful phrases as they are learned in a scrolling note-
book;

* a database of maps, museum guides and other tourist information;

A single phrasebook can be marketed in several countries since the use
of C-Level mono audio allows simultaneous coding of voice-overs in
four or five languages. Corresponding text screens can easily be provided
as well.

pea fh
pean
Rey MZ

C-LEVEL AUDIO

A five-minute picture essay on Montmartre begins with voice-over in
French, but if the tired student feels that enough vocabulary has been
learned already, this can easily be switched to another, more familiar
language without interrupting the flow of the visual sequence.

102 CD-I: A Designer's Overview

Chroma-key Facility

Students learn new words and phrases through scenes which dramatize
typical conversations, which they can control through interactive menus
and some computer control programs.

Scenes are acted out in full-motion video - and in a screen area of the
order of two-thirds by two-thirds. This is accomplished by using an
additional software decoding method and CD-I’s chroma key facility:
the actors are shot against a single color background which can be keyed
over a still DYUV background in the C-I player.

Real-time Dramatizations

The waitress approaches the customer, who asks what he can have to
drink. The student is presented with a screen menu of drink choices in
the form of an 8-bit CLUT high quality graphic which wipes onto the
screen, replacing the closeup of the waitress on the front plane. The
student can choose a drink.

Routes through the options available to the student are controlled by
computer program code which directs the branching of the operating
system to the appropriate start locations of each dialog sequence. Each
set of responses is coded onto the disc, one after the other, and each lasts
three or four seconds.

Chapter 6: Typical CD-I Applications 103

There are 3 possible sequences on the drinks menu:

° A menu of six choices appears and the student chooses café au lait

° First branch - first choice: the system moves to the beginning of the
sequence for café au lait:
Customer: ’I would like a café au lait.’
Waitress: ’A café au lait, very well;

° The waitress confirms the customer’s request and the whole menu
reappears.

(Her response is useful both to the learning exercise and to shift the angle
of view back onto her. After the student’s next interaction, the customer
can reappear at a new angle and appear to exchange remarks with the
waitress. Changing the angle of view this way avoids jarring cuts to a
new version of the same scene.)

© Second sequence - with the second menu in view, the student can re-
consider and selects jus d’ orange.
© Second branch section - all choices made after the first one: the sys-
tem moves to the start of the jus d’orange sequence:
Customer: ’No, I think I would like an orange juice after all.’
Waitress: ’Ah, you would prefer an orange juice?’

* Third sequence - this time the student confirms by re-choosing the
orange juice:
- Customer: ’Yes, please.’
Waitress: ’Very good, I’ll get it for you.’

Of course, the student could work through all six choices on the menu
without exhausting the patience of the long-suffering waitress.

104 CD-I!: A Designer's Overview

TITLE: Bienvenue 4 Paris Sheet: 8./7

@m/NTSC Normal/dbiga/Double Producer: Caeeutey
Section: Dialogues Sequence: (Cyfe ~ Drinks

PLANE A

Visual: Motion Video

Update: /$, i Sec

% screen: 76 PRY,

Customer + Waitress

Source: Chroma key - Studso

PLANE B
Visual: Stell Photograph
Update: Hold

% screen: [99 %

ae Sa ne

Source: (Commission

TRANSITION linear/branching

TIME ONSCREEN: 5 $¢¢ /module

AUDIO level: € mono/etesee

Drinks = Lip Synch.

TRANSITION EFFECT: Linear Update

2. Laterrupt to freeze

SCRIPT

Linear Nest frame
Mom Jeon Bar
Cate menu

Cate Zeon Bar

W: Bonjour ,Mansreur, vous desires
6: buest-c4e que vous aver comme horssons

Cate ambuence - S/map

Chapter 6: Typical CD-I Applications 105

CD-1 PROJECT TITLE BiNVENUE A PARIS sHeeTt: 9./9
SECTION: DIALOGUES SEQUENCE: CAFE / DRINKS

VIDEO:
General Latroduction

LEAD IN
CHOICE:

PHRASES WALLET
DIcrionaRy DIALOGUES
GUILE Boon NOTE ROOK

SeLecTe Dratooue '

CHOCCE:
DIALOGUE
CAFE STREET POST OFFICE
RAILWAY HoTec BTC.

se CAFE’

SEQUENCE ModUuLe

vioeo:
CAFE (NTRO

RETURN

CHOICE:
PRINKS MENU
& OPTIONS

ALL CHOLCES

1°*7
CHOKE APTER 13° CHOICE

VIDEO ! OURIRMATION VIDEO 2

DIALOGUE: NEW CHOICE
OfALOGUE

, q
04. ‘CAFE AU CHIT €9.'Jus D' ORANGE

VIDEO 3
DIALOGUE:
CONFIRMATION
-
PReviou £ CHorce

CURSOR
(INTERRUPT

LEAD OUT
CHOICE:

(CON BAR
MODE; CAFE SCENE

VIDEO:
MODE: ACTIVE SCREEN!’

106 CD-I: A Designer's Overview

These optional routes can all be laid out on the disc in reasonable
proximity to each other. Each section contains about six double phrases
in the exchanges between the customer and the waitress, each requiring
three or four seconds of real-time decoding.

Basic programming technique can identify a choice from one of the
action areas on the menu, then seek the first sector of the appropriate
sequence.

The customer could also ask for the bill, the option at the bottom of the
menu, which would initiate a graphic and voice-over. (The wrong bill
might be presented, which the customer could dispute by highlighting
the amount where it appears on the screen.)

The customer can pay the bill through the ’wallet’ menu on the icon bar
at the bottom of the screen, by pointing toa note ofa given denomination,
which calls up an appropriate audio response:

¢ From several denominations, the customer selects one, saying: ’Here
are .. francs’. Each possible phrase is so short that all the possibilities
can be loaded into RAM for instant response time.

The icon bar (a 3-bit CLUT graphic sub-screen) provides access to the

many different modes of operation in the lesson - a dictionary which

activates areas of the screen to provide an illustrated vocabulary, a

guidebook to Paris, even a scrolling notebook to copy the last set of

phrases for easy reference.

The attached flowchart and storyboard page show how part of this
sequence might be laid out. The storyboard page lists most of the key

Chapter 6: Typical CD-I Applications 107

factors in the design of this module such as plane configuration and
percentage of screen occupied. Transitions are indicated at the bottom
showing the next positions that the progam could move to depending
upon the user’s choices. This takes into account both linear and
interactive branching sequences.

Cop) INTERACTIVE 4)

4 Choice Sector 2 Choice Sector 3 Choice Sector

The flowchart page is a module sketch diagram used to plot the possible
actions. It forms the basis of a briefing document for the software author.
The precise routing of choices for the phrasebook application may
change several times before the final version is tested and proved
successful.

CONCLUSION

This chapter has described a range of applications as a way of further
informing the reader of specific design concepts. In these early days of
CD-I, applications will tend to be based on familiar resources - material
from other media which can be converted to CD-I, or projects modelled

108 CD-I: A Designer's Overview

on established formats from publishing, broadcast television or AV and
computer software.

CD-I is powerful because it is a multi-media publishing vehicle. All the
media can be integrated and everything is in the digital domain. CD-I is
the first medium to combine the impact of video with the power of the
computer in the same language of digital code. This demands new ways
of thinking about information storage and retrieval. This powerful new
medium will change our lives and the way in which we do things.

Chapter 7: How CD-I works 109

CHAPTER 7: HOW CD-I WORKS

For those readers wishing to understand more of the technical aspects
of CD-I after having read Chapter 5, this chapter describes the CD-I
Specification, or Green Book, in more detail. The subject is approached
from the computer-orientated aspect of the technology.

A description of the disc structure and organization is provided,
including file structure, the design of a typical CD-I decoder and a brief
description of the Compact Disc Real-Time Operating System,
CD-RTOS. Further information on CD-RTOS is given in Appendix B.

DISC STRUCTURE

The organization of a CD-I disc is designed to be compatible with
existing CD-Digital Audio (CD-DA) discs and players. CD-I is based on
the highly successful CD-DA specification and is a complete system
specification, which includes the encoding process, disc content, and the
CD-I player. CD-ROM, in contrast, only specifies a division of the disc
into sectors. The encoding and decoding methods are not defined, and
are to the individual applications developers to establish.

Disc Organization

The CD-I disc can contain some 650 Mb of data in any combination of
audio, video and computer information. All compact discs - CD-I,
CD-DA, CD-V and CD-ROM - contain a lead-in area, a program area
and a lead-out area. The program area of a CD-I or CD-DA disc can hold
up to 99 tracks, numbered from 1 to 99, and while a CD-I disc can include
CD-DA tracks, the first track must always be a CD-I track. Any CD-DA
tracks on the same disc must appear after the CD-I tracks. Typically a
CD-I disc will contain one CD-I track plus, optionally, one or more
CD-DA tracks. Each track can be of any length between 4 seconds and
the total available program space.

CD data (as opposed to music information) can be recorded in two
modes: Mode 1 - which is used in CD-ROM - contains extra error
detection and correction codes, and is suitable for data which are highly
sensitive to errors (such as computer databases). Mode 2 is suitable for
information such as audio and video data, which is not so sensitive to
errors; however, the CD-I specification has defined two specific Forms
within Mode 2, Form 1 and Form 2. Form 1 also contains extra protection
for sensitive data. All CD-I data are recorded in Mode 2.

110 CD-l: A Designer's Overview

TRACK 1 TRACK 2

166 2250 2250
Message Message Message | CD-DA
Sector Sector Sector

The track organization - which is mandatory for all CD-I discs - is
indicated in the figure. The beginning of the program area (that is, the
start of track number one), opens with 166 message sectors which contain
CD-DA information only.

The disc label (described later), comes next, followed by 2250 message
sectors (or 30 seconds), after which comes the CD-I data. Between the
end of the CD-I data and any CD-DA tracks, there must be a further 2250
or more message sectors.

Message sectors are intended to protect existing CD-DA players when
they play CD-I discs containing data which might otherwise harm the
CD-DA player or associated audio system.

Disc Label

The disc label is recorded in Mode 2, Form 1 - that is, with extra error
protection - and contains a description of all the files on the disc, its
contents, size, creator and so forth, as well as the location of any software
modules which must be loaded into the system, and the path table which
allows access to those files.

The disc label must, of course, be in Track 1 at the position shown in the
diagram above.

The disc label comprises three records: the File Structure Volume
Descriptor, the Boot Record and the Terminator Record.

Chapter 7: How CD-I works 111

Path Table and Directories

Parent Directory
Number
1

Relative Directory
Position File Name
Root
CMDS
Games
Library
Checkers
Chess
Globono
Frank
Gibson
Text
Video

“-“AOOANONAHKWN—

—_— —t
oOoOnoshsPWWWOH H+

A path table must be recorded on each disc. This provides an index of
the Directory Structure on disc typically following the disc label. Its
location on disc is given by the disc label. This is illustrated by the
example shown in the figures. The path table comprises a list of all the
directories.

Directory Structure

112 CD-l: A Designer's Overview

Each entry includes the following fields:
® Location of Directory File
© Parent Directory Name

® Directory Name
Each Directory is a file containing file descriptor records.

Files

All data on a CD-I disc are divided into files. Any file may be accessed
through the path table recorded on the disc. Each file is represented by a
File Descriptor Record contained in the appropriate Directory file which
contains the file name, number, size, address, owner, attributes,
interleave and the access permissions for read. Files may, in fact, be
interleaved on the disc so that itis not necessary for one file to end before
the new file begins.

There are several types of file: directory files, real-time files and standard
files.

CD-I SECTORS

CD-I data are divided into discrete units called sectors. These sectors are
similar (but not identical) to those specified for CD-ROM. In the case of
CD-I, these sectors contain as well as either audio, video or computer
data, the vital information which the system needs to handle this data
efficiently in real-time. (One CD-I sector is equivalent to one frame of
CD-DA.) CD-I data are recorded and transmitted at a rate of 75 sectors
per second.

CD-I Sector Format

Each CD-I sector has a total length of 2352 bytes, and apart from
synchronization information, also contains a header and sub-header,
followed by data. The header provides information on the sector address
in minutes, seconds and sectors relative, to the start of the track. It also
indicates the Mode - which in the case of CD-I, is always 2. The
sub-header comprises the following: the data type (that is, audio, video
or program-related data), the form (which may be 1 or 2), trigger bits
(including end of record and end of file) and the coding information (the
format of which depends on the data type).

Form 1 sectors contain 2048 bytes of user data and an additional 280
bytes of error detection and correction code, and are intended for data
whose integrity is essential, such as application programs, other control
data and text, where there is no built in redundancy to allow for errors.

Chapter 7: How CD-I works 113

Form 2 sectors contain fully 2324 bytes of user data, but no extra error
correction, and so are more suitable for less sensitive data such as audio
and video. Where the presence of errors will not seriously affect the
operation of the player, Form 2 sectors still allow errors to be detected
so allowing error concealment techniques to be used.

'++——— TOTAL 2352 BYTES———»

r] Form 1

Form 2

For example for video data a line or pixel in error may be replaced by
the adjacent line or pixel.

CD-I Audio Sectors

<«—____— Sector Data ——————_»

20B
\\Spare

Audio
Sector

114 CD-I: A Designer's Overview

In CD-I, audio data is held in Form 2 real-time sectors. The sub-header
contains information on the emphasis, the number of bits per sample (that
is, either 4 bits for levels B and C, or 8 bits for level A) and the sampling
rate (which is either 37.8 kHz for Levels A and B, or 18.9 kHz for Level
C). Finally, the coding information indicates whether the sector contains
mono or stereo sound.

The data in each sector comprises 2304 bytes plus a spare 20 bytes, and
is divided into 18 sound groups of 128 bytes each. Each sound group is
further subdivided into 16 bytes representing the sound parameters and
112 bytes of actual sound data. The audio data include range and filter
parameters, which are optimized in the encoding process for each sound
group.

Relative Sector Number

ra Sco
Lents SU obx ofitobstofsdofxto belo

LovelB S rloloblolabltotobetolol

: pieiobetteblolo Pope
teveic s [1000000] 0|x10] 0] o[0Jo]o]o]x|
“ Pdetoetetetsietietetetle tole

[~~] Audio Sector

The audio sectors are interleaved as shown in the figure. For real-time
audio, Level A stereo sectors are in alternate sector numbers (that is, in
the relative sector numbers 0, 2, 4, 6 ...) while at the other end of the
range, Level C mono sectors are encoded only in every sixteenth sector.

CD-I Video Sectors

CD-I Video data are also contained in real-time Form 2 sectors. The
coding information for video sectors comprises:

® the resolution (that is normal, double or high);
® the coding method (DYUV, CLUT, etc.);

® an even/odd lines flag, which is used for error concealment.
The video data is transferred directly to memory and decoded in the video
processor. Visual images are coded on disc as described below.

Chapter 7: How CD-Il works 115

DYUV Images

4 bits 4 bits 4 bits

Each pixel pair is represented by 2 bytes (16 bits) organized as shown in
the diagram. DY, DU, DV are the differential (or delta) values of
luminance and color difference. Each 4-bit value is converted to an 8-bit
value (representing 0 to 255) which is added to the previous value in the
decoder. In calculating these delta values, the encoding process must take
into account quantization errors to avoid the effect of cumulative errors.
Also, since negative values are achieved by ’wrap around’, the encoder
must avoid a pixel pair.

CLUT Images

8-bit CLUT

7-bit CLUT

4-bit CLUT

116 CD-I: A Designer's Overview

There are a number of different CLUT modes. CLUT-8 uses 8 bits of
information to define the full pallette of 256 colors.

CLUT-7 or 7-bit CLUT provides 128 colors, also at normal resolution.
For double resolution, CLUT-4 or 4-bit CLUT provides only 16 colors
but allows twice as many pixels horizontally as normal resolution, which
can be valuable for text screens particularly in complex character sets
such as Japanese where details become more visible in double resolution.

CLUT images are coded with 1 byte per pixel (normal resolution) or 1
byte per pixel pair (double resolution) as shown in the diagram.

RGB 5:5:5 Images

Upper Bank=

1bit 5bits 2 bits

Lower Bank= |

T=Transparency bit

The two banks are coded separately.

Run-Length Images

In normal resolution, 2 bytes of data are used to define first the color
(taken from the pre-selected CLUT table) and then the number of pixels
which will appear in that color before the next color change - that is, the
number of pixels for which the color will be retained. In double
resolution, pixel pairs are defined together.

Run-length images are coded as 7-bit (normal) or 3-bit (double).

Chapter 7: How CD-I works 117

7-bit Run_Length

3-bit Run_Length

Pixel Pair
3-bit
Run of Pixel Pairs

Program-Related Data Sectors

Program-related data sectors are always Form 1, since they needa higher
level of protection than simple audio and video data. Computer data in
CD-I may comprise application programs, control data, text or character
fonts, in real-time or non-real-time, depending on the application.

Real-time sectors generally contain control data and synchronization
information which is associated with audio and/or video sectors.

CD-I DECODER

CD-I decoders must be designed so that all discs can be played on all
decoders. This implies that each decoder must be designed to meet one
specification and certain parts of the decoder will be common to all.

118 CD-I: A Designer's Overview

The CD-I specification defines a base case decoder - that is, the minimum
configuration which can be called a CD-I decoder. The figure contains
a block diagram of such a decoder. It comprises:

¢ A compact disc player, plus CD-DA decoder and controller. (These
may be identical to those found in current CD-players.)

The audio processor, including ADPCM decoder and audio proces-
sing unit and attenuators.

The compact disc control unit needed to provide the random access
for CD-I and to provide real-time decoding of the sector information.

The microprocessor.
The DMA controller.

The video processor and access controller.

The Random Access Memory
Non volatile RAM.

The clock and calendar.

The X-Y pointing device.

The optional keyboard.

The Read Only memory containing the CD-RTOS operating system.

System Bus
These are described on the following page.

A number of extensions are possible to the base case decoder, including
high resolution video and peripherals such as printers and modems.

CD-Drive/Player

The CD drive allows access to any part of the CD-I disc within at most
three seconds. Drive control functions include pause, continue, stop and
eject. The CD-DA decoder and controller decodes the CD-DA data,
providing 16-bit PCM (Pulse Code Modulation) with left and right
channels to the audio processing unit. It also provides all the necessary
error correction facilities for CD-DA and low level correction facilities
for CD-I sectors, and the low level control of the CD-drive.

The CD-control unit provides selection of the required channels (that is,
channels 1 to 16 for audio, and 1 to 32 for video). It provides
de-interleaving of the audio-visual programmable data streams and
selection of file number and selective interrupt generation. The use of

Chapter 7: How CD-I works 119

CDI-BASE CASE DECODER

Left Right

Speakers

" VIDEO”
RGB|*) DISPLAY —

~ Video -
_ Decoder —

DMA hoses
FControlial _ Controller —

F Cloc 1
_ Calendar —

~ ROM
" (CD-RTOS)

SYSTEM
BUS

120 CD-I: A Designer's Overview

the sub-header in each sector is very important to select the data type and
to de-interleave the audio, video and program related data.

PCM
Decoding

CDI
Interface

Video &

Data out Decoding LL

System Bus

It should be noted that the CD-player with the CD-DA decoder and
controller is capable of playing a normal CD-DA disc as well as CD-I
discs.

The CD-DA controller decoder outputs CD-DA PCM data to the audio
processing unit, and the CD-control unit outputs CD-I data, which can
be either ADPCM, (Adaptive Delta PCM) audio, video data or program
related data.

The CD-control unit recognizes the type of sector which is being read by
decoding the sub-header and manages it accordingly - for example,
ADPCM audio data may be sent directly to the audio processor or to the
memory, while video and program-related data will be sent to the
memory.

The de-interleaving of data ensures that the video, audio and program
related data are sent to the appropriate parts of memory.

Audio Processor

The audio processor comprises the ADPCM decoder and audio
processing unit. ADPCM data is passed to the ADPCM decoder either
directly from the CD-control unit or from memory. The ADPCM decoder
decodes Level A, B or C audio.

This outputs PCM audio data, both left and right channels.

Chapter 7: How CD-I works 121

CD-DA
PCM Data
Special
Effects
Processor

Speakers

ADPCM
Decoder

The Audio Processing Unit

All audio passes through the Audio Processing Unit (APU) on its way to
the playback system. The APU acts like a simple two-channel audio
mixer within the CD-I system, with two input channels, either stereo or
mono, from two sources, and can mix the input channels to the left and
right output channels under the control of the application.

In addition, the APU can vary the loudness of each channel separately,
for mixing and speaker panning effects. Individual units of sound, stored
on disc, can be used in various ways for specific application software.
Without the APU function, all audio combinations for a specific
application must be pre-authored and stored on the disc in case it is called
up by the user. A great deal of redundant storage space is saved this way.

Video Processor

The video processor includes the access controller and video decoder.
The access controller provides an interface between the main system bus,
the memory and the video decoder. The random access memory is
divided into two banks: Bank 0 and Bank 1. Bank 0 contains information
for Plane A, and Bank 1 for Plane B. The memory is also used for storing
soundmaps and for system memory.

The video decoder provides all the facilities for decoding the various
pictures and special visual effects described in Chapter 3. The schematic
diagram of the video decoder is shown in the figure. It comprises two
real-time decoders, each capable of decoding visual images coded as
described elsewhere in this chapter.

122 CD-I: A Designer's Overview

External Video
(Optional)

Video
Decoder RGB

RAM | RAM
Bank 0: Bank 1

Real-time decoder 0 decodes DYUV, CLUT and Run-length images and
real-time decoder 1 decodes DYUV, CLUT, Run-length and RGB
images. RGB images, of course, require the use of both memory banks
and provide only one full screen image per plane.

Plane A

Plane B

The output from each decoder is 8 bits each of Red, Green and Blue. For
DYUV, CLUT and Run-length the output of each real-time decoder is
at least 6 bits per R, G and B. RGB 5:5:5 provides only the most

Chapter 7: How CD-I works 123

significant 5 bits of each component. The three least significant bits are
always zero. The two planes A and B may be overlayed in either order
(i.e. A in front of B or B in front of A) and means are provided such as
the color keying and mattes for creating transparent areas in each plane.
Alternatively, the planes may be mixed and the brightness of each plane
defined by multiplying by an image contribution factor which allows for,
e.g. fading. The pixel hold function is provided between the overlay and
the mixer as shown. (Pixel repeat for CLUT and RGB 5:5:5 is provided
within each of their decoders.)

Dyuv Decoding

DY (4 bits)
Y (8 bits)

Interpolate

Intermediate

(8 bits| Pixel Values
each)

Backdrop/
External

YUV Start Values _ Video

The backdrop and cursor are added at the last stage where the digital
signals are also converted to analog Red, Green and Blue. For the base
case decoder, the analog signals must have at least 64 levels (equivalent
to 6 bits of digital information). Optionally, the decoder may provide the
full 8 bits at 256 levels for each R, G and B.

The real-time decoder for DY UV is shown in the figure. The basic coding
unit for DYUV is a pixel pair requiring two successive bytes in memory.
For each pixel pair there are two values for Delta Y, one value for Delta
U and one value for Delta V. In the decoder, for each of the Y, U and V
signals, the delta codes are converted to their non linear quantization
values; that is values between 0 and 255 (see figure) and the successive
delta values are summed to give the absolute Y, U and V values. For each
line, if not black, the absolute start values for Y, U and V must be
provided and coded separately on the disc. U and V are linearly
interpolated to fill in the intermediate pixels, because only every other
pixel is sampled. Finally, Y, U and V are matrixed to RGB for displaying.

124 CD-Il: A Designer's Overview

Chapter 3: Run-length
Coding

For RGB 5:5:5, each pixel is represented by 16 bits; 5 bits for each of
the R, G and B components and one bit to provide transparency. This
requires the use of both banks of memory to provide only a single image
plane. RGB 5:5:5 decoding is, therefore, very trivial.

For CLUT coding, we have 1 byte representing 1 pixel for CLUT 7 and
CLUT 8, and 1 byte representing 2 double resolution pixels for CLUT 3
and CLUT 4. CLUT 7 is available in both image planes, whereas CLUT
8 is only available for plane A.

For Run-length images, Run-length 7 uses the 7 bit CLUT and a run is
represented by 2 bytes; the first byte giving the CLUT address and the
second byte giving the length of the run which can be up to 255 pixels
long. For CLUT 3 (which allows double resolution) the first byte
provides the colors for a pixel pair and the second byte gives the
Run-length. For both 7 and 3 bit Run-length coding, all lines end witha
zero length run, which means that the color should be continued to the
end of the line.

| __]Allowed | _]Not allowed

Chapter 7: How CD-I works 125

To use CLUT coding, whether directly or Run-length coded, it is
necessary to know which entries in the CLUT are actually used. There
is one color look up table in the decoder which has 256 entries, so for 8
bit coding the whole of this CLUT is allocated to plane A. (Note: It is
not possible to use the CLUT for both planes A and B since this would
mean time sharing it. Speed constraints prohibit this.) For 7 bit coding
the CLUT is divided into two halves, giving 128 entries for plane A and
128 for plane B. Therefore, itis possible to have simultaneous 7 bit CLUT
images. 4 bit CLUT coding uses the last 16 entries in each half and
Run-length 3 uses the last 8 entries in each half. An application program
can load appropriate color values into each entry of the CLUT, as
appropriate.

The figure indicates the allowable combinations of the different coding
methods. There are only two real restrictions on combining images. The
first is thata CLUT 8 image cannot be combined with any other CLUT
image or Run-length, as it uses the whole CLUT (all 256 entries). It
therefore can only be used with DYUV in plane B or, of course, with
plane B off. The second restriction is that RGB 5:5:5 uses the data stream
from both image memories, so it cannot be overlaid with any other digital
video plane.

RGB levels

As mentioned before, the output from each real-time decoder is an RGB
signal converted into analog with a color resolution of at least 6 bits per
R, Gand B component. This means that each componentcan havea value
from 0 to 255. The CCIR have recommended, for TV studio operation
involving digital signals, that black level should be represented by 16 on
a scale from 0 to 255 and peak white by 235.

This allows some space below 16 and above 235 for inevitable
undershoot and overshoot which is caused by bandwidth limitations of

126 CD-l: A Designer's Overview

the TV signal chain. (For example, a square wave signal would overshoot
both white and black levels when the bandwidth is limited.)

To avoid distorting images, the CD-I specification follows these CCIR
recommendations. These are indicated in the figure. These levels are
particularly important for natural images encoded with DYUV but to
maintain consistency, they are used for all image coding methods. The
RGB levels must be taken into consideration in the encoding of visual
images.

Display Control Program (DCP)

The control of visual images include the loading of CLUT values, the
defining of mattes, etc. The CD-I specification defines what is known as
the display control program, or DCP. This consists of a set of display
control instructions which are carried out on every field scan of the
display. These instructions are placed in two tables which are the field
control table, which is executed once every field, and the line control
table, which has separate instructions for each line of the display.

_<< 1 Display Line

: Start Matte A
: Change ICF

: Start Matte B
: End Matte A
: Change ICF
End Matte B

mOoadcown

Matte B
ICF Area

Matte A

There is a field control table and line control table for each of two image
planes. The field control table can hold up to 1024 instructions and is
used for setting up the display parameters at the start of each field and
can include the data for loading into the color look up table for example.
The line control table comprises up to 8 instructions for each line. These
allow the display parameters to be changed from one line to the next as
the scan progresses.

Chapter 7: How CD-I works 127

One of the most basic uses of the line control table is to provide a display
start address for the beginning of the image line. It is possible, therefore,
for the start of each line of an image to be in random locations in memory.
This is particularly important for providing scrolling of images,
particularly together with the use of subscreens, where only part of the
screen is scrolling, where there will be a discontinuity between the fixed
part and the scrolling part of the screen.

One important application of the DCP is the creation of mattes. An
example is shown in the diagram. Here the display comprises two
overlapping mattes and what is known as an ICF (image contribution
factor) area. If we look at one line of the display marked in the figure,
there are six points on this line where there is a boundary of one of thése
areas. Matte A is circular, matte B is a diamond and so on, but each of
these mattes or areas, can in fact, be almost any shape. Along this display
line point a indicates the start of matte A; b is the start of the ICF area,
at which point, the ICF is changed to a new value. At point c is the start
of matte B; d the end of matte A; e the end of the ICF area and therefore
a change back to the old value of ICF and f marks the end of matte B.

Scanline Position 1
(Rabbit held in position) ~~

Scanline Position 2
(Rabbit free to move)

Signal when this line reached, to
avoid flicker when redrawing.

Matte A and matte B could, for example, represent areas of trans-
parency/translucency, revealing the plane beneath and can be used to
generate some of the visual effects which were described in Chapter 3.
Since the positions a, b, c, etc. can be changed from line to line, there
can be a large number of mattes on one picture. The limitations are that
only two mattes at most can overlap and there can be, at most, 8 matte
boundaries on any one display line. The figure, of course, indicates six
such boundaries.

128 CD-I: A Designer's Overview

Since each line of the LCT comprises up to eight instructions, it would
be possible to change all of these boundary values on each line. However,
this would leave no room for other instructions which may be required,
for example, to set a new line start position.

Another example of a DCP instruction is shown in the diagram. This
provides synchronization to display scanning. In the figure the example
shown is an object being moved to a new position, for example, to
produce animation.

To avoid flicker in the signal it is important that the object should not be
moved while the object itself is being scanned. Therefore, it is possible
on any line to insert an instruction which will generate a signal when that
line has been reached. Therefore, the application can wait for the signal
and can then move the object to the new position ready for the next
scanning field.

Further examples of DCP instructions are as follows: the image coding
method can be changed at any point on the display. This is how
subscreens are produced.

A new backdrop color can be loaded. CLUT colors can be changed as
the scan progresses to extend the number of colors available in any one
image. A CLUT image plane can also be changed from 256 to over 2000
colors.

The Hardware Cursor

The small hardware cursor plane, 16 pixels square, was described in
Chapter 3. Each pixel can be set to be either transparent or the cursor
color which may be defined from a range of 16 colors. The cursor may
be appear solid on the screen, or may blink, and it can be in normal,
double or (as an extension) high resolution.

The position of the cursor can be set to any position within the full screen
area of the display.

Border Color
The border around the reduced screen area can be set differently for each
line and for each plane; however, if DYUV coding is used, this border
must be black.

Chapter 7: How CD-I works 129

DMA Controller

DMA
Controller

The DMA controller allows fast, efficient transfer of data within the
system, between the CD control unit and memory, and from one part of
memory to another, without the use of the MPU, thus releasing the MPU
for more important work. Its function is shown in the figure.

Random Access Memory (RAM)

i)
Access
Controller

RAM | RAM
Bank 0: Bank 1

512 kB 512 kB

The random access memory (RAM) comprises two banks of 512k bytes
each. It is accessed through the access controller using the
micro-processor unit and the video processor. Each bank comprises 256k
words, 16 bits wide. Image data has to be stored in the appropriate bank

130 CD-I: A Designer's Overview

(Plane A or Plane B); other data, such as soundmaps and system data,
may be stored in either bank.

Non-volatile RAM (NV-RAM)

A non-volatile random access memory (NV-RAM) is also included
within the CD-I decoder. It comprises a minimum of 8K memory, either
non-volatile or backed up by a small battery to provide continuous
support in the event of a power failure. Part of the NV-RAM is reserved
for system information. The remainder may be used for application data.
Data is organized as files in the NV-RAM.

Clock/Calendar

A battery-operated clock calendar, accurate to at least one minute in a
month, records hours, minutes, seconds, the month, year and leap year.

Pointing Device

One or more X/Y pointing devices are provided as user interface for each
CD-I decoder. Each will also include two trigger buttons and must be
capable of pointing at any pixel on the full screen display at normal
resolution.

The pointing device may be a mouse, joystick, lightpen, tracker-ball or
any other suitable device, giving either absolute or (in the case of a
mouse) incremental position, or velocity of movement.

Transfer Paths

At this stage it is worth a review of the material covered in this chapter
up to this point by looking at the various pathways that data can follow
within the CD-I system.

Data moves through the CD-I system from storage on disc or in RAM,
from synthesis in the Microprocessing Unit (MPU), or from screen
actions by the user. Application software tells CD-RTOS when and
where to find data and which pathway to send it along.

Application Software
Application software is taken from the disc when the program begins and
loaded into RAM to direct the program. Software can be loaded in and
out of RAM during the course of the program depending on variable
routes chosen by the user. This data is mixed with the audio and video
data in the data stream.

Transfer Paths
CD-DA

Control
Program(s)
(Application

Software)

DM
P

Video
Processor

Plane B

Backdrop

Data

---- Control

Chapter 7: How CD-I works 131

132 CD-l: A Designer's Overview

Audio Pathways

Audio can travel along four different routes depending on its source and
use. CD-DA is played direct from the disc through the Audio Processing
Unit (APU) to the playback system and does not enter the CD-I
controller.

Most CD-I audio at A, B or C-Level is played directly from the disc in
real-time. Playback is controlled by the CD control unit, which channels
digital audio data to the APU for playback.

Audio from the disc is the determining factor in thereal-time recovery of
all data read from the disc. While audio is being read from one location,
the player cannot look elsewhere for other data such as a picture. Thus,
sound and pictures required in the same sequence must be interleaved in
the data stream. To have an audio sequence with a range of choice for
accompanying pictures, the audio must be stored in multiple locations
on the disc, interleaved with a new set of pictures at each location. This,
in turn, would require a lower audio quality or a slower rate of picture
change.

Playback from System Memory

Soundmaps can be loaded into temporary memory from disc or the MPU.
Like direct playback from disc, digital sound data is converted to analog
signals in the decoder and played through the APU.

Video Pathways

~ VIDEO ~
DECODER

“VipEO
DECODER |

All video images are loaded into RAM before display on the video
screen. Sectors of video are interleaved with sectors of audio in the data

Chapter 7: How CD-l works 133

stream. Video data can be retrieved from the CD disc or synthesized in
the MPU before being stored in one of the two Image Stores (RAM).

Image Store (RAM)

From RAM, the video data passes through a decoder which sends it to
the appropriate screen plane for display on a monitor or television. The
video decoder has two real-time decoders which convert digital video to
analog RGB signals in the appropriate resolution. One real-time decoder
receives imagery from Image Store 0 (RAM 0) through path 0 which,
when decoded, will be passed to the 8-bit image plane supplied by that
decoder. The other decoder is supplied by Image Store 1 via Path 1.

The real-time decoders control screen effects by designating the screen
area and the contribution of each screen to the overall image.

A dissolve is a simple example of this: a one-second dissolve from a
desert scene on Plane A to a camel caravan on Plane B requires the
controller to increase the full screen contribution of Plane B from 0 to
100% over one second as it decreases the contribution of Plane A from
100% to 0.

CD-RTOS

At the heart of every computer is its operating system, the program
controls all the other software programs and hardware. CD-I has its own
Compact Disc-Real-Time Operating System (CD-RTOS), designed
specifically for demands of interactive multi-media information
management in real-time. It is based on a high-performance operating
system called OS-9, which is concisely written in 68000 assembler code.
This section summarizes the main features of CD-RTOS. Further details
are given in Appendix C.

Organization

The organization of CD-RTOS is shown in the diagram. At the top is the
application program which communicates with the CD-RTOS kernel
through various libraries. Housekeeping modules are connected to the
kernel, and below the kernel are the file managers: the Compact Disc
File Manager, Pipe File Manager, Non-volatile RAM File Manager, and
the User Communications Manager.

Below the managers are the various drivers, which interface directly with
the hardware described above. The drivers allow for differences between
CD-I players and configurations from different manufacturers.

134 CD-l: A Designer's Overview

Chapter 7: How CD-| works 135

THE KERNEL

The kernel is the heart of CD-RTOS. It provides multi-tasking, handling
all task control itself, including task switching and task suspension. It
also handles memory management (including the allocation and
de-allocation of memory as the application demands), all the interrupts
generated in the CD-I player, all the systems service requests, the I/O
(input/output) calls and so on which pass through it.

Finally, it handles control between individual tasks and synchronization,
a vital task. There are several ways of communicating between tasks:
simple software interrupts that the processes can send between each
other; events or semaphores; shared memory in the form of data modules;
or pipes which allow communication through the I/O system.

Configuration Status Descriptor

The configuration status descriptor (CSD) is an integral part of the CD-I
system. Basically, it allows an application to discover what devices are
available on the player and information about these devices, such as their
capability. There is an entry in the CSD for each device on the system,
stored in RAM/ROM with additional parts in non-volatile RAM. It is
also possible for the user to configure the machine using the CSD
mechanism, perhaps to indicate a choice of control device.

Each CSD entry, known as a device status descriptor, includes four parts:
the device type, name, active status (such as whether it is busy), and a
set of parameters which are specific for each device.

All the devices in the decoder, including the base case devices, will be
represented in the CSD, which will indicate, for example: the capabilities
of the video processor (for example, whether it can do high resolution),
the audio processor, the NVRAM size and so forth - all of which will be
described in the configuration status descriptor.

The CSD also provides information for any extensions or peripherals
which may be connected to the decoder - floppy disc, keyboard, printer,
modem and so on.

Start-up Procedure

When the CD-I decoder is switched on and a disc inserted, the software
performs the start-up procedure. The hardware is initialized, the
CD-RTOS kernel is started and the system copyright message is
displayed. After that, the boot file is loaded from the disc, or, if there is
no boot file, the configuration status descriptor is compiled.

136 CD-I: A Designer's Overview

STARTUP PROCEDURES

Chapter 7: How CD-l works 137

The file system is then initialized. This involves reading the Disc Label
and the Path Table which shows the layout of what is on the disc. After
that, it may display the manufacturer’s label and copyright message,
before executing the first application program pointed to by the disc
label.

If the user removes the disc and inserts another, this process is repeated
from the point indicated in the diagram.

File Protection

The disc file protection mechanism restricts access to certain files: this
allows a content provider to protect up to 32 different files or
combination of files per disc through access codes. This means a disc
can be designed to contain several application programmes, access to
which can be individually controlled through separate access codes and
charges: the customer may buy the disc and pay for one or two
applications, with the option of accessing others, for a fee, at a later date.

Each CD-I player has a unique code, in addition to which there is an
encryption method in the kernel which will match the player and the
access codes to control access to the files.

For example, a disc may contain up to 32 games. When the user purchases
the disc he pays a fee for this first game only. Later, he may pay for other
games on the same disc and receive an access code which is fed into the
CD-I player using the keyboard or pointer. The application must provide
the mechanism for carrying this out. (Note that the access code will be
different for each application and each player.) In this way the user pays
only for the games he wishes to play.

FILE MANAGERS

The file managers are located between the kernel and the drivers and are
used to provide the I/O despatching. The main CD-RTOS file managers
are: the User Communication Manager (UCM), the Compact Disc File
Manager (CDFM) and the Non-Volatile RAM File Manager (NRF).

UCM Video

The User Communications Manager provides software support for the
video graphics devices available, the pointer devices, the optional
keyboard and audio processor. The UCM video functions support the
organization of the two image planes, and the creation and manipulation
of multiple images in memory, and provide basic graphics drawing and
text functions.

138 CD-l: A Designer's Overview

DRAWMAPS: A drawmap is simply an area of memory which can be
used by the application to store and manipulate images, either loaded
from disc or created using the drawing functions. Drawmaps can have
several formats: DYUV and CLUT images are stored as a rectangular
array while Run-length images are stored as a list. RGB 5:5:5 requires
two drawmaps, one in each memory bank.

Examples of UCM drawmap functions include:

* ’create and close’ (used for allocating and de-allocating memory for
drawmaps);

® *transparent copy and exchange’, similar to copy and exchange’, ex-
cept that pixels of specific value are not transferred (also useful for
animation purposes).

If a drawmap is larger than full screen size, the display may be scrolled
around the larger image. Drawmaps smaller than the screen size may be
used to create animation effects: for example, partial updates may be
handled as small drawmaps, a succession of which are overwritten in turn
onto a larger drawmap. The use of transparency allows a non-rectangular
object to be overwritten on a larger drawmap.

GRAPHICS DRAWING: The basic graphics drawing functions are line,
rectangle, polyline, circle, ellipse, arc and fills. In addition, text in almost
any font style and character shape can be drawn. The graphics drawing
parameters include the use of patterns, definition of colors, line size, dash
style, fonts, writing modes and clipping region.

Other functions include OR, AND and exclusive OR. In practice these
functions operate on multi-colored patterns and drawmaps, encoded as
either CLUT or RGB.

REGIONS: Regions may be used to limit the drawing area within
drawmaps. In addition, the data which is generated by the regions can be
used in the display control program to create mattes for special effects.

Regions can be created in several basic shapes: rectangles, polygons,
wedges, ellipses, and so forth. In addition, these basic shapes can be
combined through region operations. Regions may also be used to clip
drawmaps: so, for example, that part of a line which lies outside the
region can be clipped.

Chapter 7: How CD-I works 139

DISPLAY CONTROL PROGRAM: The UCM provides functions to
load the display control program which was described above.

UCM Audio

The audio part of the user communications manager controls the flow of
ADPCM sound and control signals from the MPU or memory to the
ADPCM decoder. It supports the creation from the application, specific
sounds and allows the sounds to be manipulated in memory as well as to
support the hardware mixing, panning and attenuation capabilities of the
audio processing unit. It manipulates soundmaps, which are basically a
memory area that an application can use to preload data from the disc.
Soundmap manipulation functions allow different types of audio data to
be mixed together before they are sent to the audio processor.

UCM User Interface

The UCM also supports the X/Y pointer devices and optional keyboard,
and provides functions for the application to read data from the keyboard
and the position indicated by the X/Y pointer device.

Compact Disc File Manager (CDFM)

The Compact Disc File Manager converts high-level commands from
the application to device driver commands. It provides access to and
interpretation of the disc file system, carries out any necessary access
protection and schedules the disc access.

It may be that the application which is running requires multi-tasking, or
that more than one application is running on the system: the CDFM
schedules these accesses to the disc so that access requests are processed
in the order in which they appear on the disc - which may not be the order
in which they appear in time.

This process works something like a lift, which from the ground floor to
the top, stopping not in the order in which the passengers indicated their
individual floors, but as the floors themselves appear.

The following service requests are processed by CDFM:

* Change User’s Default Directories

* Change Current File Position Pointer (Seek)

® Read Data From File

* Get Specified Status Information

140 CD-I: A Designer's Overview

¢ Set Status Information
© Perform Special Function

* Close a File

Note that Seek allows movement within a data file, real-time record or
real-time file, to skip over certain real-time records in the program.

NV-RAM File Manager (NRF)

The NV-RAM File Manager provides a filing system for the application
to write files to the NV-RAM, through a simple version of the Random
Block file manager available on OS-9 for magnetic discs. It provides a
basic directory structure with only one NV-RAM directory and no
sub-directories. File names may be generated by the application,
providing there is no duplication.

Since the amount of memory is limited (a minimum of 8K bytes for the
base case decoder), the application must use this memory very
efficiently.

DRIVERS

The device drivers provide the necessary interface between the file
managers and the devices themselves. The drivers will take account of
any differences in devices from different manufacturers so that a standard
interface is presented to the manager and therefore to the application.
The drivers will handle all of the low-level functions which the file
managers provide to the application.

SYNCHRONIZATION AND CONTROL

Synchronization, which co-ordinates audio and video effects in the final
presentation, can be achieved in several ways:

¢ A software timer, which can be instructed to generate a timed signal.
¢ A user input, which can also generate a signal.
¢ A signal from another application, which is running concurrently.

° A trigger bit, an end of record bit or end of file bits in the sub-header
of sectors.

Chapter 7: How CD-I works 141

Control of the flow of data through the system is achieved using the Play
Control Structure, while synchronization can be achieved using the
trigger settings in sectors and the Real-Time Control Area, both
described below.

Play Control Structure

Real-time files typically contain interleaved sectors of audio, video and
program-related data. The Play System Call used to play real-time files
selects sectors by file number, channel number and data type (audio,
video or program-related). It also controls the destination of the data
through the Play Control Blocks and Play Control Lists, and signals the
application software whenever a specific event occurs in the program.

At the heart of Play System Call is the Play Control Block, which
provides the means to de-interleave the data as it arrives in the player,
and store audio, video and program related-data in the appropriate parts
of the memory. The Play Control Block points to the audio, video and
data Play Control Lists.

Audio data may, of course, be sent directly to the audio processor or to
the memory as asoundmap. The Play Control Block provides the channel
selection mask to indicate which audio sectors are to be placed in
memory and which should be directed to the audio output.

The Audio Play Control List points to the soundmap to which the audio
data should be directed. Similarly, the Video Play Control List points to
the appropriate drawmap, and the Data Play Control List to the area
where the data is to be directed in memory.

Real-Time Control Area

The Data Play Control List buffer includes text and control data, the latter
usually ina Real-Time Control Area (RTCA) which is interpreted by the
Real-Time Record Interpreter (RTRI). This is used to control the
real-time playback and synchronization of a real-time file.

There is a Real-Time Control Area at the beginning of each real-time
record; the RTRI decodes the function codes within the RTCA for both
the audio and video effects and provides synchronization information.

The Real-Time Control Area contains a number of commands which are
used to control the playing of real-time records, the loading of audio and
video data, the manipulation and display of drawmaps, and the output of
audio data.

142 CD-I: A Designer's Overview

It is also used to synchronize audio and video, and contains facilities for
user interaction. The commands in the Real-Time Control Area can be
executed in sequence or in parallel or, more likely, a combination of the
two.

The RTRI is a truly multi-tasking interpreter which can handle a large
number of parallel tasks. This is necessary to ensure that data can be
loaded from disc and displayed for output to the audio processor while
inputting user responses.

The RTCA is a genuine real-time programming facility that provides all
the structures with which programmers are familiar, and is also capable
of multi-tasking.

InVision

A series of efficient and easy-to-use tools are a useful addition to any
language designed to be used by application programmers. An
object-orientated multi-media user interface called InVision was been
developed for CD-RTOS. While alternative User Interfaces and
associated programming tools will no doubt emerge, InVision is the first
of its kind.

InVision is designed for the programmer in the development of CD-I
applications as well as the actual user of a CD-I player. The user will
normally be presented by a version of the InVision user interface
modules adapted to the specific CD-I disc by the programming team who
developed the application.

However, InVision is more than a User Interface. It contains a series of
tools and functions designed to integrate into CD-RTOS. These collected
into three main modules:

© Visual Shell: the only part of InVision that the user will actually see,
this essentially forms the control panel of the CD-I player.

Presentation Support Library: a collection of sub-routines intended to
simplify the task of manipulating the display, and obtaining input from
the user.

Display Manager: a CD-RTOS sub-routine module for the UCM file
manager which provides access to the video, audio, pointer and key-
board driver routines.

Further details of InVision can be found in Appendix C.

Appendix A: Technical Specification Summary 143

APPENDIX A: TECHNICAL SPECIFICATION
SUMMARY

The following technical specification information is provided for your
information, and has been used for the calculations and other technical
statements made in this book.

Units of Measurement
1 Mega byte = 1,024 kilo bytes
1 Kilo byte = 1,024 bytes

1 Sector = 2,352 bytes with headers and synch info.
= 2,048 bytes (Form 1) user data
= 2,324 bytes (Form 2) user data
= 2,336 bytes (CD-DA)

Disc Capacity

650M bytes

Disc Transfer Rate

75 sectors/sec = 150k bytes/sec (Form 1)
= 170.2k bytes/sec (Form 2)
= 171.1k bytes/sec (CD-DA)

N.B. Above is useful data rate, less headers, etc.

Screen Resolution

NTSC: 360 x 240 pixels = 86,400/1024 = 84.4k bytes
PAL : 384 x 280 pixels = 107,520/1024 = 105k bytes

Typical Values
1 DYUV natural image picture = 85k bytes NTSC (105k bytes PAL)
Partial update of 1/3 by 1/3 of screen (= 1/9 of screen area)

= approx. 8.5k bytes = 10%

So with a data rate of 170.2 at 15 frames per second you can have
170.2/15 = 11.35k bytes per frame, based on either partial update of a
DYUV natural image picture, or a full screen update of RL-7 CLUT
cartoon picture

N.B. Partial update can be improved to some 50% of the screen area by
software techniques.

144 CD-I: A Designer's Overview

1 second CD-DA sound = 171.1k bytes

1 second Level A Stereo = 85.1k bytes
Mono = 42.5k bytes

1 second Level B Stereo = 42.5k bytes
Mono = 21.3k bytes

1 second Level C Stereo = 21.3k bytes
Mono = 10.6k bytes

Max theorectical playing time = 74 mins 33 secs
practical playing time (CD-DA & form2 CD-I = 72 mins
practical playing time (CD-ROM & form1 CD-I) =65 mins

Disc design must take the following factors into account:
® disc storage capacity

® system RAM allocation

® microprocessor load
The whole disc can contain:

7,830 DYUV NTSC natural image pictures (650M bytes = 650x 1024/85)
6,340 DYUV PAL natural image pictures (650M bytes = 650x1024/105)

or
65 minutes of cartoon animation plus Level C mono sound

17 hours 18 mins of Level C mono sound (16 channels x 65 mins)
or

100 million words (150,000 pages of text)

Appendix B: Glossary of terms

APPENDIX B: GLOSSARY OF TERMS

absolute disc addresses The location of a given sector on the disc in
minutes, seconds and sectors, contained in the header of the sector.

absolute RGB coding See direct RGB coding

absolute RGB components Every color can be represented as the sum
of different proportions of the three primary colors, red, (R), green (G)
and blue (B). In absolute RGB encoding, every pixel is represented by
its R, G and B components. These are values which, on presentation to
suitable digital to analog converters will give the correct voltages
required by the red, green and blue guns of a cathode ray tube to produce
the color of the pixel on the display screen.

absolute sector address The address part of the sector header field. Its
value corresponds to the absolute disc address.

absolute time In CD-DA, the total time a disc has been playing. Included
in the subcode and thus available for display during playback.

access The process of locating information in a data store.

access controller A CD-I player component that takes drawmaps from
RAM and loads them to the video decoder.

access key The key supplied by a content provider to allow access to a
group of files protected by a file protection code.

access protection The method of preventing unauthorised access to
specific data of a confidential nature, stored on a disc.

active line scan period The time taken by the electron beam of a cathode
ray tube to move across the visible part of a line on the screen.

active display The contents of a video memory currently being
displayed, as opposed to screen contents being held in memory for later
display if needed.

active regions Areas on a display screen which respond when indicated
by a pointing device/cursor under user control. Used for user input in
interactive systems such as CD-I.

145

146 CD-I: A Designer's Overview

adaptive delta pulse code modulation A technique for converting
analog audio into digital audio in any CD-I level. Delta modulation
assumes close correlation between successive samples. It cannot
accurately express large transients in an audio signal, because the
correlation between successive samples is too low. Adaptive delta pulse
code modulation is a variant of delta modulation in which the
quantization steps are adapted to the dynamic amplitude variation. This
adaptation can include a temporary switch to PCM. See delta modulation,
pulse code modulation.

ADPCM See adaptive delta pulse code modulation.

address of path table The block address of the first block of the system
path table.

ADPCM decoder Device which converts CD-I audio sector encoded
data to 2’s Complement 16-bit PCM encoded audio.

album A set of discs.

American Standard Code for Information Interchange (ASCII) A
standard data transmission code designed to achieve compatibility
between data devices.

analog versus digital In analog systems, natural sound and images are
converted into corresponding variations in electrical voltages or currents.
The resulting electronic signals vary according to the original sound or
image in a completely linear fashion. These signals aree then processed
in various ways, such as making and playing a record or tape, or
transmitting the signal for reception at a remote receiver. The signal is
then amplified and reproduced locally. The same lineare technique was
used in the past in the telephone system, for transferring the spoken voice
from one telephone to another.

One of the major disadvantages of analog technology is the occurence
of noise and distortion. Noise occurs in the form of disturbances to the
original signal caused by extraneous effects in the electronic circuitry
used in the analog system. Distortion arises when any part of the signal
processing system fails to represent the original signal accurately. Noise
and distortion in an analog system are virtually impossible to remove, as
they aree superimposed on the signal information itself.

The concepts behind digital systems were first developed for telephony
systems, where increasing demand for long distance communications
broughtattention to the need to reduce or eliminate noise and interference
signals. The development of the semiconductor, and with it the mass

Appendix B: Glossary of terms

produced computer components now commonplace in the area of
professional electronics, brought digital technology into the consumer
area.

Digital circuitry works on the principle of defining all signals as a range
of specific values or numbers, rather than by a continually variable
voltage or current. In digital recording, the original signal is captured,
and converted into a sequence of values, each representing the
instantaneous level of the signal at a given moment. By repeating these
samples at a fast enough frequency, or sampling rate, all the changes in
the original sound or picture can be accurately recorded.

animation The art or process of synthesizing apparent mobility of in-
animate objects or drawings. See CLUT animation.

appliance controller A dedicated circuit, between a computer and an
associated appliance, through which the computer controls the appliance.

applications software A computer program written for a specific user
application.

ASCII See American Standard Code for Information Interchange.
aspect ratio The width to height ratio of an image or a pixel.

audio block A block of audio information in CD-I format.

audio block field bytes 2304 bytes of data in a CD-I audio sector. The
audio block is further subdivided into 18 sound groups of 128 bytes each.

The sound groups have to be encoded in sequential order.

audio channel Audio data from one source. Up to 16 audio channels
can be encoded in a CD-I track.

audio data (1) Audio information expressed in digital form. (2) In
CD-DA, multiplexed and pulse code modulated stereo information with
CIRC and subcode added. (3) In CD-I, audio information encoded in
accordance with the CD-I specification.

audio functions User Communications Manager functions which are
concerned with the maintenance and manipulation of ADPCM sound.

audio mixing control unit Adjusts the volume and balance distribution
of audio information.

147

148 CD-I: A Designer's Overview

audio processing unit Converts digitally coded audio information to
the ’left’ and ’right’ analog outputs. Also includes an Audio Mixing
Control Unit.

audio sector A data field containing audio data.

audio sector data format The data field of an audio sector comprises
a sub-header, an audio data block of 2304 bytes and 20 bytes with value
0.

audio quality level See sound quality level.

audio track A CD-DA track with information encoded as 16-bit wide
2’s-Complement numbers. A separately addressable section of a
CD-DA disc, normally carrying a self-contained piece of music. Has a
minimum duration of 4 seconds and a maximum duration of 72 minutes.
One CD-DA disc can contain between one and 99 audio tracks, but the
total disc playing time cannot exceed 72 minutes.

authoring The work involved in producing the software for a CD-I
application, from the initial concept to the recording of the master tape
required for the disc mastering process. Authoring embraces:

Encoding the required audio, video, text and binary data into CD-I
data formats.

Developing and producing the application software which operates
on, uses or accesses the encoded CD-I data as required by the appli-
cation.

Structuring the encoded CD-I data and application software into the
disc label, files and records corresponding to the access and playback
requirements of the application.

Verifying and validating the application so produced via, at the least,
a CD-I disc/base case combination.

authoring process In CD-I, the process of developing and producing the
complete software for an application. It involves (a) Designing the
program content by creating the story board. (b) Creating and capturing
data, and preparing it for use. (c) Developing the program that will appear
on the disc. (d) Simulating and testing the program in practice. (e)
Preparing the final master tape.

The first step in the process is the creation of an overall program design,
or story board. This is critical to the success of the project, as it defines
not only the type of basic material or data required for the video, audio

Appendix B: Glossary of terms

and text/graphics components, but also the interrelationships between
them. The concept of interactivity, in which the response of the system
depends upon the response of the user at each stage of the program,
means that the whole project stands or falls upon the ’what happens then’
response of the overall CD-I system. This involves both the player and
the disc, as well as the user. The key to this action-and-response
relationship must be clearly defined in the story board design phase if
the program is to be sucessful.

It is unlikely that the story board will get the design right at the first
attempt. One of the essentials of CD-I program design is step-by-step
evaluation of results, leading to redesign and re-evaluation in an
on-going interactive process. With both the application data and the
application software on the disc, (all the information on a single
medium),

There is no opportunity to correct errors once the disc is released for
replication and sale.

So simulation, evaluation, validation, testing and, where needed,
revision, are essential steps in every phase of the CD-I authoring process.
And the controlling mechanism for this process is the story board.

Based on a first issue of the story board, design work proceeds along two
parallel paths. The first involves content collection; the assembly and
production, when needed, of the video, audio, text and graphics
information required by the story board. This basic information may be
in either digital or analog form, or may have to be generated from scratch.
Continuing along this first path, the collected data is tested and evaluated
at each phase for correctness, and for the right balance of quality versus
disc capacity. While Compact Disc-Digital Audio sound takes 100% of
the information channel from the disc to the player, monaural speech
takes only 6% of this capacity, leaving 94% for other information - video,
text/graphics, application software or indexing information. The choice
of a balance between quality and data bandwidth is a key element in
program design.

Once satisfactory, data is encoded and compressed to CD-I format, and
prepared as data files onto a disc simulator. This is built.around a large
capacity read/write store (typically hard discs are used, with a capacity
in excess of 1200 megabytes) and is part of the authoring studio
equipment).

The second path involves the development and testing of the application
software and user interfaces needed to interact with the data used in the
program.

149

150 CD-I: A Designer's Overview

Each stage of this application software development is testd on its own,
and slowly integrated and synchronized with the relevant data as it in
turn is collected.

In this way, the integration and synchronization of each portion of the
program is tested to prove it works as defined in the storyboard, before
integration into the next level of the design.

And once completed the first time, the program is then tested, revised,
and retested on the disc simulator until the overall design and balance
has been proven.

Only at this stage can thought be given to transferring the total program
to the disc replication facility.

authoring system In CD-I, a general term for the equipment, in hardware
and software, needed for authoring CD-I discs.

authoring tool A computer programming aid used in authoring.

auto-play, auto-play mode In CD-I, the play sequence of a CD-I disc
when no user input is given, or when the CD-I player is so instructed by
the user. Used for demonstration or training of the disc application, or
for general information sequences.

auxiliary data field InCD-ROM and CD-I the last 288 bytes of a sector,
either used for extra error detection and correction (mode 1 and form 1)
or available as user data area (mode 2 and form 2). See mode 1, mode
2, form 1, form 2.

backdrop The background image plane that is displayed when all other
planes are made transparent.

background plane Synonymous with backdrop.

base case disc In CD-I, a hypothetical disc that can exercise all the
capabilities of a Base Case system. See Base Case system.

base case See Base Case system.

base case system The minimum characteristics of a system that may
bear the CD-I name.

BER See bit error rate.
bit error rate A measure of the capacity of a data medium to store or

transmit bits without errors. Expressed as the average number of bits the
medium can handle with only one bit in error. CD-ROM and CD-I, which

Appendix B: Glossary of terms

employ three layers of error detection and correction (CIRC and
EDC/ECC) have a bit error rate of 10°” (one error per 10'8 bits).

bit inversion A random error causing erroneous read-out of a bit, a 1
becoming 0, and vice versa.

bit-mapped display A screen display in which each pixel location
corresponds to a unique memory location whose contents determine the
intensity and color values of the pixel.

bit map The process by which a picture is built up on a pixel-by-pixel
basis.

bit mapped text Specific characters generated by bit mapping.

bits per sample The number of bits used to express the numerical value
of a digitized sample.

block In CD-ROM and CD-I, the user-data portion of a sector.

block address A 32-bit integer that is converted to an absolute disc
address to access information on the disc.

boot file Optional File on a CD-I disc containing a program to be
executed when the CD-I disc is first mounted. It can be used to add or
replace Operating System modules in the Base Case system.

boot identifier Identifies the operating system which supports this boot.

boot record An optional record on a CD-I disc that specifies where the
Boot File is on the disc. The contents of the Boot File is loaded into
memory when the disc is first mounted.

border The area outside a visual image which is reduced below the full
screen size.

burst error The corruption of a sequence of bits caused e.g. by a read
error, tracking error or electromagnetic interference.

CARIN CAR Information and Navigation system developed by Philips
for computerized on-the-road route planning and route following, using
digital maps recorded on CD-ROM or CD-I discs, in association with
navigation sensors.

151

152 CD-l: A Designer's Overview

cartesian coordinate system A system for locating a point in a plane
by specifying its distance from two axes which intersect at right angles.

cartoon-style image An image containing significant areas of the same
color which can be efficiently Run-Length coded.

CAV See constant angular velocity.
CD See Compact Disc.
CD-DA See Compact Disc Digital Audio.

CD-DA controller/decoder The hardware needed to play a CD-DA or
CD-I disc and to decode the information coming from the disc, either in
the form of CD-DA audio information or other information passed to the
digital output port of the controller/ decoder.

CD-DA data Data encoded according to the CD-DA specification.

CD-DA track A track on a Compact Disc containing music encoded
according to the CD-Digital Audio specification.

CD device driver The lowest software level to handle CD drives. The
only software to communicate directly with the CD control unit, it resides
in ROM ona CD-I player.

CD-disc master A CD master disc, produced by exposing a
photosensitive coating on a glass substrate to a laser beam. The laser is
modulated by the digital program information from the CD-tape master,
together with the subcode, which is generated during the disc mastering
process from the subcode cue code, also on the CD tape master. The
exposed coating is developed, covered with a silver coating and nickel
plated to form a ’metal father’ recording mould. See CD Mastering,
Metal Father.

CD drive The drive mechanism portion of any CD player.

CDFM See Compact Disc file manager.

CD graphics In CD-DA, a technique for generating text, still pictures or
animated graphics, related to the music. The graphic information is
recorded in subcode channels R-W. Presently used in Japan only. Not

related to the graphics facilities of CD-I.

CD-I See Compact Disc-Interactive.

Appendix B: Glossary of terms

CD-I channel The main channel of a CD-I track corresponding to the
specifications of CD-ROM, mode 2 and the CD-I logical and physical
formats.

CD-I disc design See authoring process.

CD-Interactive digital audio In CD-I, there is a requirement to have
audio data on disc at a number of distinct quality levels. These quality
levels are equivalent to LP record quality, FMradio quality, AM radio
quality, and telephone quality. Further, CD-Digital Audio information
can also be played on CD-I equipment.

In addition to CD-DA sound in 16-bit pulse code modulation (PCM)
format, CD-I audio data is also coded in 8- or 4-bit Adaptive Delta Pulse
Code Modulation (ADPCM) formats. This technique is chosen as a way
of coding sound more efficiently than for CD-DA, such that 50% or less
of the total data rate is occupied by stereo audio information. At least
50% of the data rate can therefore be used for other purposes, principally
the transfer of visual information. The Hi Fi music mode uses an 8-bit
word size and a sampling rate of 37.8 kHz in order to take full advantage
of the form 2 sector space of 2324 Bytes, while remaining the highest
usable integral fraction of 44.1 kHz (i.e. the 16-bit PCM sampling rate).
Hi Fi music mode is equivalent in quality to a high-quality LP played for
the first time. In order to use the same coding technique to span the
requirement for various audio levels and still maintain optimal quality
by proper post-filtering, the word size of the first level is reduced from
8 bits to 4 bits to give the Mid Fi music mode. This is equivalent to FM
broadcast quality sound as broadcast from the studio, and offers a
maximum of 4 stereo or 8 mono channels in parallel as opposed to the 2
stereo or 4 mono channels available in the Hi Fi music mode. To achieve
a further reduction in data rate, and thereby increase the number of audio
channels to 8 stereo or 16 mono, the sampling rate is reduced by half to
18.9 kHz. This results in the speech mode quality, which is equivalent
to AM broadcast quality sound as broadcast from the studio.

It should be noted that a channel as described above is equivalent to some
70 minutes of uninterrupted playing time. Multiple channels can only be
played with a 1-4 second gap between them. This gap is due to the fact
that the laser read-out mechanism must be repositioned back to the
beginning of the disc.

An alternative way of using the channels is as a sequence of up to 16
parallel channels of audio information. These channels could tell the
same story but in different languages, for example, so that the user could
switch from one language to another instantly at any time. This last case
moves away from the question of what is on the disc alone, to the question
of how that information can be used in a CD-I system.

153

154 CD-I: A Designer's Overview

Audio information from the disc can reach the user in three different
ways. (1) From the disc directly to the 16-bit PCM decoder, and out
through the audio Hi Fi system as CD-DA sound. (2) From the disc
directly through the ADPCM and PCM decoders and the Hi Fi system
as ADPCM sounds. (3) From disc into a microprocessor-controlled
random access memory, where it can either be held awaiting its singular
or repeated use whenever a certain event occurs (for example a ball
bouncing on the screen, which must be accompanied by the appropriate
sound), or it can be slightly altered as a function of different events and
then sent under microprocessor control through the ADPCM and PCM
decoders and out to the Hi Fi system. This latter approach allows for
audio interactivity with a quality that has been unachievable in the past.
See adaptive delta pulse code modulation, delta modulation, pulse code
modulation.

CD-Interactive digital video In CD-I, there is a requirement for various
video quality levels to offer a choice of resolution and color depths to
satisfy various pictorial functions in the applications. Three resolution
levels are defined: the best achievable resolution for pictures on present
normal TV receivers (normal resolution); the best achievable resolution
for characters displayed on present normal TV receivers (double
resolution); the best achievable resolution with the coming
enhanced-quality TV sets (high resolution). As for color depth, the
quality necessary depends on the type of image that is being handled.
Natural stills use YUV (luminance and color signals) coding for an
equivalent of 24-bit total color depth per pixel, quality graphics employ
Color Look-Up Tables (CLUTs), and user-manipulated graphics use
direct RGB coding.

A key requirementis that the disc must be compatible regardless of where
it is purchased and on which system is is used, i.e. playback should be
independent of the particular TV standard. Given these and other
similarities and differences, CD-I video requirements are translated into
specifications related to three areas:

¢ display resolution

* picture coding

* visual effects.

CD-I systems will work with, and CD-I disc contents will be displayable
on, normal TV sets. The video coding adopted conforms, as far as
possible, with prevailing industry conventions relating to color depth,

visual effects and studio world considerations, while remaining
independent of the TV standard (525/625 lines).

Appendix B: Glossary of terms

The starting point for the specification of resolution, in addition to
525/625-line display systems compatibility is to ensure the readability
of text, the undistorted shape of graphics, and the full screen view of
natural pictures on display. To do this, two sets of resolution areas are
defined; one as a safety area for text and graphics, the other a full screen
for natural and animated pictures. Moreover, three disc formats are
defined:

® a525/625-line-compatible format that can be used in the international
market to satisfy all compatibility requirements.

Each format is usable on each display system with, however, a quality
penalty. The basic numbers for normal resolution, i.e. the best resolution
visible on a non-interlaced TV, are 384 x 280 pixels for full screen and
320 x 210 pixels for the safety area. For maximum readability of
characters on a normal TV display, the double resolution mode is
defined.

This mode has twice the number of horizontal pixels as the normal
resolution mode. For future programs, but still keeping in line with the
data rate limitations of Compact Discs, a high-resolution mode is defined
as twice the horizontal and twice the vertical resolution of the normal
resolution mode, giving 768 x 560 pixels for full screen and 640 x 420
pixels for the safety area. This is also consistent with high-resolution or
525/625-line-compatible digital TV.

The net distortion is at most 7% for a 525 or 625-line disc on a 625 or
525-line decoder, respectively, and 3.6% for compatible discs on 525 or
625-line decoders.

This is aconsiderable improvement on the 20% distortion obtained when
NTSC (525 line) material is transferred to PAL (625 line) systems for
viewing. Also, it relates quite favourably to the fact that the eye can only
resolve, at best, distortions of 5% if the original and the distorted object
are side-by-side on the same screen; if they are not side-by-side the eye
can only resolve, at best, distortions of 10% for objects of a familiar shape
(e.g. circle, square etc). Even these low CD-I distortions will be reduced
to zero when real-time pixel manipulation is added to CD-I equipment.

155

156 CD-I: A Designer's Overview

As far as picture coding is concerned, three target areas are defined by
CD-I applications:

* natural pictures

® animation

In all cases, it is necessary to use compression techniques to decrease the
amount of data required for a given picture, and thus the loading time
and memory storage requirements. By using compression, the
achievable update speed, the number of images that can be put on a disc,
and the number of video ’channels’ available per unit are all increased.
Clearly, video coding needs to be simple so that the decoding can be
affordable, and it must also be capable of being performed in real time
while still maintaining image quality.

The compression techniques chosen are:
© 1-dimensional delta YUV (DYUV) for natural pictures
© direct RGB coding for high-quality end- user manipulated graphics

® Color Look-Up Table (CLUT) for graphics and fast update and mani-
pulation

Each of these techniques gives optimal performance in the area of use
for which they were chosen. In particular, with DYUV coding there is
no visible difference between the compressed and original pictures. Four
different sizes of CLUT, two used with Run-Length coding, offer an
efficient trade-off between the number of colors needed and the rate of
change required. If pictures coded in DYUV or CLUT graphics are
intermingled with audio data in speech mode quality, for example to
explain their content or to enhance a story, then 3 full-screen
pictures(normal resolution mode compatible format) can be displayed
every 2 seconds while the audio is playing. Moreover, for full-screen
animations like cartoons, real motion is achievable with the Run-Length
compression coding specified.

The data rate of the CD-I data channel is not high enough for visual
effects such as cuts and wipes to be performed in the full motion video
data stream (as is done in movies, for example). Furthermore, for
interactive use it is desirable to have not only more channel space (for

Appendix B: Glossary of terms

more effective use of CD-I), but also the ability to vary the visual effects
used on the same picture data as a function of end-user activity or
computer software state. Visual effects are therefore approached at a
higher level, e.g. via control functions in the data stream, rather than
embedding them uniquely in a moving data stream.

As far as operations on a single visual plane are concerned, CD-I at the
basic system level will be capable at least of:
* cuts

* smooth x,y scrolling

efficient updating of any part of a visual field independent of the con-
tents of the rest of the visual field

trading-off picture resolution against visual data thoughput, for a con-
stant visual field size.

The overlaying of images is based on the ability to have at least one
hardware cursor plane available, one, two or three independent
full-screen full-picture visual planes available, and one backdrop plane
available for use with external video at a pixel level. These plane
combinations allow CD-I to be used for a variety of applications as
encountered in multi-cell film work, e.g. in animations, in games-like
manipulations of objects over objects, or with a background over a
foreground. In CD-I, the control of overlays is done by a transparency
bit for the pixels of the cursor, and also the RGB plane. As for the use of
the CLUT, acolor key is used to control the overlay of such planes, while
for DYUV the pixel-wise overlaying of regions under well-defined
transparency/translucency conditions is used.

The final point concerning CD-I visual is that of operations between two
visual planes. There are rwo main categories of such operations:

® wipes, and

® dissolves or fades.

Wipes and dissolves are well known in the film and video industries as
well as, in a simpler form, in professional slide shows. They are very
important in maintaining an attractive presentation potential in CD-I for
both stills and moving picture sequences. These effects, together with
the other visual effects described, bring CD-I as close as possible to the
present passive video world, while at the same time allowing for
interactivity via either end-user or software control of visual information.

157

158 CD-I: A Designer's Overview

This multi-faceted control of visual content and visual information flow
as perceived and influenced by the end-user makes CD-I potentially the
richest artistic medium ever created. See color look-up table, delta YUV,
direct RGB coding and Run-Length coding.

CD-I physical format The way in which data is stored on a CD-I disc.

CD-I physical sector Directly addressable sector in a CD-I disc
numbered consecutively beginning at zero.

CD-Isector A unit of data of 2352 bytes.
CD-I system A real-time system capable of playing CD-I discs.

CD-I track A data track containing only mode 2 sectors conforming to
the CD-I specification.

CD-master tape A master tape for Compact Disc, containing all the
program information in the required digital format, and organized in the
correct relationship. See CD mastering. Compare CD-tape master.

CD-mastering In gramophone record production, mastering is the
process of recording the information from a master tape on to the master
disc. This same basic procedure applies to all Compact Discs
(CD-Digital Audio, CD- Video, CD-ROM and CD-I) but with two major
differences:

* an additional ’pre-mastering’ stage is required to arrange and encode
the information in the required Compact Disc format, including CIRC
error detection and correction coding, subcode insertion and eight-to-
fourteen modulation.

the recording is ’cut’ by exposing a rotating master disc with a photo-
sensitive coating to a laser beam modulated by the signal from the CD
Tape Master, as it traverses the disc.

In keeping with the microscopic dimensions associated with CD
recording, the photosensitive coating is only one-tenth of a micron thick.
It is deposited by centrifuging on an optically ground and polished glass
disc to form a ’resist master’.

After exposure, controlled photographic development produces a pattern
of pits in the photoresist. To capture this pattern, and thus create a CD
Glass Master Disc, silver is evaporated onto the pit pattern.

From the glass master, replication moulds are made along conventional
lines by nickel plating and discs are fabricated by injection or
compression moulding. But the microscopic dimensions and the

Appendix B: Glossary of terms

requirement of optical read-out have demanded new materials, processes
and techniques for stringent quality control.

CD-ROM See Compact Disc-Read-Only Memory.
CD-RTOS See Compact-Disc Real-time Operating System.

CD-RTOS kernel The nucleus of CD-RTOS which is responsible for
service request processing, memory management, system initialization,
multitasking, input/output management and exception and interrupt
processing.

CD-tape master The tape used to produce the CD-disc master; a CD
master tape with subcode cue code added. See CD mastering, cue code.

CD track A separately-addressable section of a Compact Disc, normally
carrying a self-contained piece of information.

CD-Video Optical recording system extending the CD-DA standard
with analog video. Discs may be 12 cm CDV singles, 20 cm EP
(extended play) or 30 cm LP (long play) sizes. CD-Video is compatible
with the earlier LaserVision standard, so that CD- Video players can play
LaserVision discs, of both CLV and CAV types.

CDV See CD-Video.

CD-Video player Device specifically designed to play CD-Video and
CD-DA discs. If provided with a digital output and control interface,
can also be used in conjunction with a suitable signal processor, to read
CD-ROM or CD-I discs. See also omni player.

CD-V single 12 cm CD-Video disc carrying up to 6 minutes of video
with digital stereo or 2-channel sound, plus 20 minutes of CD-DA sound.

channel Data blocks within a real-time record can be labelled according
to logical ’channels’ which can be selected in real time. It is by selecting
these channels that the user can change audio, video and text sources
during the playing of a real-time record.

channelnumber A number assigned to pieces of information contained
in the real-time record to facilitate the selection of such information.

chroma key See color key.

159

160 CD-I: A Designer's Overview

chrominance Information on hue and saturation (vividness) of an
image. Expressed as U and V signals. Compare luminance.

CIRC See Cross-Interleaved Reed-Solomon Code.

clipping In computer graphics, to remove parts of a display that lie
outside a selected area. In signal processing, a restriction on the peak
amplitude of a waveform.

CLUT See color look-up table

CLUT animation In CD-I, a technique used to impart motion to graphic
objects by repeatedly changing the data in the color look-up table. For
CLUT coded images, the 256 values in the CLUT control the colors of
the entire image, therefore, some simple animation effects are possible
simply by redefining some or all of the CLUT contents as a function of
time.

CLV See constant linear velocity.

color look-up table A table containing all the colors which may be used
ina particular picture. Each entry is an absolute RGB value. The picture
may then be encoded using the table addresses rather than the absolute
values.

combi player A CD-Video player.

Compact Disc System for reproduction of high-density digital data from
an optical disc. Originally conceived as a medium for high-fidelity
music reproduction, for which Compact Disc-Digital Audio is now an
accepted world standard. Because of the very high disc data storage
capacity, Compact Disc is now being applied as a text/data medium for
electronic publishing (CD-ROM) and a multiple-function (audio/
video/text/data) medium for interactive programs (CD-I). See Compact
Disc-Digital Audio, Compact Disc-Interactive and Compact Disc-Read
Only Memory.

Compact Disc-Digital Audio Developed jointly by Philips and Sony,
and launched in October 1982, Compact Disc-Digital Audio has
revolutionized high fidelity recording with its pure sound reproduction,
small size and immunity from surface scratching.

The Compact Disc system records music, in the form of digital data, onto
a light but robust 12 cm diameter disc, thereby virtually eliminating the
problems of dynamic range, background noise, wow and flutter, and

Appendix B: Glossary of terms

other sound disturbances common to earlier sound recording systems.
32-bit analog-to-digital conversion at a sampling rate of 44.1 kHz., in
conjunction with CIRC (Cross Interleaved Reed-Solomon Code) error
correction and EFM (Eight-to-Fourteen Modulation) a reproducable
bandwidth of 10 Hz to 20 kHz within 0.2 dB, a signal to noise ratio of
over 100 dB, a signal to noise ratio of over 100 dB, a dynamic range of
over 95 dB and imperceptible wow and flutter.

But the feature that distinguishes Compact Disc from other audio
recording systems is the fact that it is based on optical recording
technology. Optical recording was invented in the Philips Company’s
research laboratories in Eindhoven, Holland in the late 1960s, and has
formed the basis of a range of optical discs products.

The principle behind optical recording is the use of a small laser to burn
minute pits in an optically flat surface which is enclosed in a transparent
sandwich disc construction. For playback the recorded surface is
illuminated with a lower power laser. The light beam is concentrated onto
the protected recording surface as the disc is rotated. Light is reflected
from the recording surface, and passed to a photo sensor. The amount of
light reflected from the disc surface changes depending on whether or
not the beam is passing over a hole or pit made during the recording
process.

The pits, between 0,9 micron and 3.3 micron long and 0.6 micron wide,
are recorded on a spiral track at a pitch of 1.6 micron, sixty times finer
than the pitch of an LP record. The track is approximately 5 km long.
The music recorded in a normal recording studio is encoded into the
CD-DA format and used to drive a recording laser to produce a master
disc. Mechanical stamper discs are produced from the original master,
and copy discs are then pressed in quantity from these stampers in a
specially developed replication process.

The high recording density achievable with optical recording techniques
results in over one hour of high quality sound recorded on one side of
the disc, the other side being used only for the disc label. CD-DA discs
are designed for playback at a constant linear velocity. This means that
the speed at which the track is scanned by the laser pickup is a constant
1.25 metres/sec. As a result, the speed of rotation of the disc changes as
the disc is played. Play starts at the inside of the recording track, and runs
outwards. As the music is played, the rotational speed of the disc drops
from some 500 rpm at the beginning of the disc, to some 200 rps at the
end. Apart from the main data channel via which the hi fi music is stored,
an additional 8 sub-channels, with a much lower data capacity are also
available for control and display purposes. These sub-channels, known
as P,Q and R through W, are generally available to the recording studio.
However, apart from the P and Q sub-channel, little use is made of these
sub-channels in practice. While the R through W channels have in fact

161

162 CD-I: A Designer's Overview

been specified for a simple graphics application, no sub-channel graphics
decoders have as yet become generally available on the market. The disc
surface is divided into three main portions. The lead-in area at the centre
or start of the disc, the program area and the lead-out area at the outside
of the disc. In the lead-in area, the Q sub-channel is used to store details
of the contents of the disc. Up to 99 separate music tracks can be
specified. A single track has a minimum duration of 4 seconds, and a
maximum duration of 72 minutes (the whole disc). The location of each
track on a given disc in terms of absolute time in minutes and seconds
relative to the start of the program area of the disc, as well as the running
time of each track, is defined in the Q sub-channel in the lead-in area.
The P sub-channel carries a music flag for quick track finding using a
simple decoder.

The Compact Disc-Digital Audio Specification, also known as the Red
Book, is available to paid-up licensees. This specification contains full
details of the CD-DA system. See also CD mastering, Cross-Interleaved
Reed-Solomon Code, Eight-to-Fourteen Modulation, Pulse Code
Modulation.

Compact Disc drive Device specifically designed to read digital data
from CD-ROM or CD-I discs. CD-I drives can also play CD-DA discs.

Compact Disc file manager A software module which handles I/O
requests for the compact disc drive. Provides random access to files at
the byte level through system calls.

Compact Disc-Interactive The Compact Disc-Interactive standard
specifies a multi-media, interactive information carrier that is mainly
real-time audio- and video-driven, but also has text, binary data and
computer program capabilities. It is both a media and a system
specification, and defines what can be present on the disc, how it is coded
and organized, and how disc/system compatibility can be maintained.

Multi-application-based CD-I is targeted at the consumer electronic and
institutional markets. It aims at satisfying a wide range of application
demands for both these markets.

From a technical point of view, CD-I is based on CD-ROM, but from a
player/product point of view it is based on CD-DA. Like CD-DA, it is
dependent on processor hardware, but unlike CD-DA or CD-ROM, it is
also system-software dependent. The reasons for CD-I’s hardware and
system-software dependence are motivated by, and based on, the
real-time audio/video decoding and data-handling requirements that
CD-I applications demand, as well as the requirement to maintain
disc/system interchangeability in the same way that CD-DA does. In

Appendix B: Glossary of terms

practical terms, this means that any CD-I disc will be able to be played
on any player, regardless of where in the world both were purchased.
This latter point is achieved in the CD-I specification by defining a set
of rules for a minimum level system called the Base Case, which must
be observed by all discs.

The CD-I specification also allows for mixing of CD-DA and CD-I tracks
on CD-I discs, and requires CD-DA decoding hardware in CD-I systems.

The CD-I specification is a complete standard that: (1)is applicable to
the consumer market; (2) can be realized as a one-disc, interactive,
multi-media content carrier (i.e. a CD-I disc) by various content
providers (e.g. publishers, the audio-visual industry etc.); (3) is capable
of being produced by the existing CD manufacturing facilities; (4)
assures disc/system compatibility and is, for this reason, resistant to
system variations.

Satisfying all these customer requirements, and in so doing giving still
higher performance levels for selected application areas, forms the basis
for CD-I as an interactive multi-media carrier.

Present and potential future CD-I applicatons can be categorized as
follows:

°
°
°
°

Entertainment

adventure games

activity simulation

‘edutainment’

oo Oo 80

Creative leisure

° drawing/painting
° filming

° composing

Work at home/while travelling
© document processing
° information retrieval and analysis

163

164 CD-I: A Designer's Overview

Compact Disc player Device specifically designed to read CD-DA
discs. If provided with digital output and control interface, can also be
used, in conjunction with a suitable signal processor, to read CD-ROM
or CD-I discs.

Compact Disc-Read-Only Memory A natural derivative of Compact
Disc-Digital Audio. Defined by Philips and Sony in 1985, the CD-ROM
makes use of the identical physical characteristics - disc size, rotational
speed and read-out mechanism, as well as the same disc mastering and
replication processes as used for CD-Digital Audio.

Where CD-ROM and CD-DA differ is in their application. Instead of a
single, dedicated application namely hi fi music, the CD-ROM
specification limits itself to defining the method by which data is stored
on the disc, and no more. The nature of the data, and the purpose for
which it is to be interpreted, is left to the information providers making
use of the medium. The disc can be divided into tracks in the same manner
as for CD-DA; indeed the specification foresees the possibility of
combining CD-DA tracks with CD-ROM tracks on a single disc.

CD-ROM makes use of the same CIRC error protection used in CD-DA
as well as EFM (Eight-to-Fourteen Modulation). However, the data
recorded on the disc is organized into sectors of 2352 bytes. Each sector
is further subdivided. After 12 bytes of synchronization, and a 4-byte
header to identify the address and nature, or mode, of the data in the
block, the main User Data area follows, containing 2048 bytes of data.
Following this area is a 288 byte long Auxiliary Data Area.
Concerning the mode information, CD-ROM normally only uses mode
1, where an additional level of error protection (EDC/ECC) is included
in the Auxiliary Data Area to reduce the chance of error to less than a
single bit per disc. Mode 2, also defined for CD-ROM, allocates the space
used in mode 1 for error correction for recording additional user data.
Mode 0 is used for CD-DA applications. The mode being used is fixed
for the duration of a track. Details of the modes of all tracks are also held
in the Q sub-channel in the lead-in area of each disc.

The data rate from the disc is 175 kilobytes per second. This means
175/2.352 = 75 sectors per second. The minimum length of each track is
4 seconds, and a track contains a minimum of 300 sectors, each sector
containing one 2048 byte block of user data in mode 1 and one 2336 byte

Appendix B: Glossary of terms

block of user data in mode 2. The total disc can contain 72 minutes of
data, or in mode 1, (72 minutes) x (60 seconds) x (75 sectors) x (1 block)
x (2048 bytes) = 663.5 megabytes of user information. This increases to
756.8 megabytes for mode 2. A typical type A4 page contains some 4000
bytes. Thus a CD-ROM mode 1 disc could contain over 165,000 pages
of typed text. Of course this does not allow room for the necessary
indexes to enable the data to be searched.

There is a selection of CD-ROM drives on the market. A series of
standard hardware interfaces are also available to connect CD-ROM
drives to personal computers.

CD-ROM is finding a well-defined place in the professional world for
the distribution of bulk databases. Typical applications replace text based
microfiche publishing, or on-line databases. In order to locate data on
such a large-capacity disc as CD-ROM, personal computer versions of
mainframe database retrieval software are normally used. Such software,
adapted to the requirements of CD-ROM drives, used in conjunction with
inverted files to identify the specific occurrence of each given word in
the total database, can be used very effectively. Typical seek times of 3
to 10 seconds are now quite normal for text-based data bases of several
hundred megabytes.

A further feature of CD-ROM is that the data is stored in digital form,
and as such data retrieved from the disc can be reprocessed or re-edited
by suitable word-processing software.

During 1985 and 1986 an ad-hoc group of CD-ROM information
providers and other related companies met to attempt to extend the basic
CD-ROM specification to cover such matters as file structure, file
directory index, and operating system. This group, known as the High
Sierra Group after the hotel in Lake Tahoe where the group first met, has
now completed its study, and recommendations have been passed to the
appropriate standards committees (NISO and ECMA). While Philips and
Sony as owners of the basic CD-ROM standard remained aware of the
work of the High Sierra Group, and indeed provided comments and
suggestions to the group, no attempt was made to update the basic
specification with this additional input from the High Sierra Group.

Compact Disc-Real-time Operating System CD-RTOS, the operating
system used in CD-I, is specified so that the real-time capabilities of CD-I
are usable, as far as possible, in a device-independent way. The features
of CD-RTOS are that it:

* isa multi-tasking operating system with real-time response, has a ver-
satile modular design, and can be loaded into ROM.

* supports a variety of arithmetic and I/O co-processors

165

166 CD-I: A Designer's Overview

® is device-independent and interrupt-driven
* can handle multi-level tree-structured disc directories

® supports both byte-addressable random-access files and real- time fi-
les and

* is OS-9 compatible.

CD-RTOS is composed of four major blocks:

Libraries; these guarantee that the necessary specialized user library
functions such as high-level access and data synchronization, as well
as math, I/O and other functions are available in CD-I systems. One
of the most important of these is synchronization.

* CD-RTOS kernel; this is a customized version of the OS-9 kernel.

Managers; these define the virtual device level for graphics, visuals,
text, audio, CD control etc. The managers provide software support
for graphics/visual devices, pointing devices, and the CD-I audio pro-
cessing devices, as well as taking care of disc I/O and optimized disc
access and reading.

© Drivers; these are the interfaces between the virtual, i.e. hardware-in-
dependent, level and the actual hardware used by various manufactu-
rers in their CD-I systems.

compatibility In Compact Disc, the extent to which different types of
discs can be interpreted by different types of players or drives. For
example, all CD-DA discs are fully compatible with all CD-DA players,
so that any player can reproduce music from any disc regardless of
manufacturer.

compression A technique in which the amount of information used to
present a specific image is reduced by eliminating redundant or
unnecessary information.

computer graphics Pictures created by computer programs. Standard
drawing functions such as ’line’, ’circle’, etc. are normally used.

concealment In digital signal processing, the hiding of errors e.g. by an
interpolation scheme.

concurrent (audio) channel Block multiplexed audio information.
CD-I allows for up to 16 audio channnels to be recorded concurrently on
the disc, so that by playing the disc several times and accessing different
channels, extended playing time is obtained or by playing the disc once,

Appendix B: Glossary of terms

the same or related audio information may be obtained from parallel
channels, e.g. the same information in a different language.

configuration status descriptor Describes the configuration of a
particular CD-I system. Composed of device status descriptors.

constant angular velocity A disc rotation mode in which the disc always
rotates at the same speed, so that the time of one revolution is always the
same.

constant linear velocity A disc rotation mode in which the discrotation
speed changes as the read radius changes so that the linear reading speed
i.e. the speed at which the read-out device scans the track, is always the
same. Maximizes disc information storage capacity.

content provider In CD-I, the writer, publisher or other party(ies)
supplying information, usually copyrighted, for an application.

CRC See cyclic redundancy check.

Cross-Interleaved Reed-Solomon Code (CIRC) An error protection
code specially developed for Compact Disc. It consists of two
Reed-Solomon codes interleaved crosswise.

CIRC makes it possible for a CD player decoder to detect and correct or
conceal large burst errors. Errors up to 4000 data bits (2.5 mm of track)
can be corrected. Errors up to 12,304 data bits can be concealed.

The CIRC encoder uses 2 stages of encoding and 3 stages of interleaving.
The 12 PCM audio samples (24 symbols) of one Compact Disc audio
frame are fed in parallel to the first encoder. The second symbol of each
audio sample is delayed by two symbols, so that the symbols of two
successive frames are interleaved. The first encoder then adds 4 parity
symbols, making 28 in all.

These. 28 symbols are fed to the second encoder through delay lines of
different lengths. The second encoder adds four more parity symbols,
making 32 in all. Finally, alternative audio signals are delayed by one
symbol. The total effect is to spread the symbols of one frame over eight
frames.

The two stages of CIRC encoding make it possible for the CIRC decoder
in a CD player to correct two symbols in each received frame directly,
or to correct four symbols in each received frame by erasure and
calculation. Furthermore, it allows detection of up to 32 successive
incorrect symbols so that interpolated values can be substituted. Because
of the dispersion of symbols over 8 frames, up to 4000 wrong data bits

167

168 CD-I: A Designer's Overview

can be corrected and up to 12,304 wrong data bits can be concealed. The
final (1 symbol) delay provides protection against random errors. See
also Compact Disc-Digital Audio, eight-to-fourteen modulation.

CSD see configuration status descriptor

cue code In Compact Disc, a code used in tape mastering. Recorded on
audio track 1 of the CD-tape master, it contains the information necessary
to generate subcode during disc mastering. See CD mastering.

cursor plane A small graphical image plane that can be moved around
the display or made invisible as required. It can be positioned at any
position over the other planes.

cut A basic effect in film and video editing which causes an image to
appear, usually to replace a previous image.

cut and paste An electronic technique for the manipulation of textual or
pictorial information on a display screen in a manner similar to the cut
and paste technique used in editing such information on paper.

CVBS Composite video broadcast signal. The standard form of color
TV broadcast signal in which the intensity and relation of the red, green
and blue components are represented by a luminance signal and a
chrominance signal.

cyclic redundancy check In Compact Disc, a separate error detection
scheme for the Compact Disc subcode.

DAT See Digital Audio Tape.

data channel (1) In CD, a channel carrying data, as opposed to audio
information. (2) In CD-ROM, a channel carrying mode 1 data.

data driven action tagging In CD-I, the technique for identifying or
tagging events on the different data streams (audio, video, text/data) so
that they can be synchronized according to the requirements of the
application program.

data integrity The preservation, against loss or corruption, of programs
or data for their intended purposes.

data symbol An entity made up of n bits representing a value between
Oand 2".

Appendix B: Glossary of terms

data track A track with information encoded as 8-bit wide symbols
(bytes) organized in sectors.

delta modulation In data communications, a form of differential PCM
in which only 1 bit for each sample is used.

delta pulse code modulation See delta modulation.

delta-YUV A high-efficiency image-coding scheme for natural pictures
used in CD-I. The delta coding takes advantage of the fact that there is
a high correlation between adjacent pixel values, making it possible to
encode only the differences between the absolute YU or Y V pixel values.
This coding scheme is applied per line i.e. in one dimension. See YUV
encoding.

differential PCM In data communications, a version of pulse code
modulation in which the difference in value between a sample and the
previous sample is encoded. Fewer bits are thus required for
transmission than under PCM. In CD-I, this technique is applied in video
encoding as well as audio encoding. See adaptive delta pulse code
modulation, delta YUV.

digital audio tape Internationally agreed standard for digital audio tape
recording. Digital Audio Tape cassettes, although noticeably smaller
than Compact Cassettes, can record up to 2 hours of continuous digital
quality sound. Recordings are made diagonally on tape by a twin rotary
recording head. In a similar way to CD-DA discs, digital audio tape
recordings incorporate CIRC error protection for high sound quality and
long life. They also include a control and display subcode for highly
convenient playback.

digital magnetic tape Magnetic tape with a thin base layer for precision
tape-head contact, and a high linear density (approximately 20 times that
of analog tape) to accommodate the bit packing requirement of 20 kbits
per inch.

digital recording Recording audio or video signals in digital form. The
level of the signal to be recorded is sampled at a rate at least double the
highest frequency to be reproduced, and the instantaneous amplitude of
the signal is quantized and stored in numerical or digital form.

direct RGB coding Picture coding scheme used in CD-I for high-quality
(e.g. modelled) graphics that can easily be changed by the user. Images
are encoded on disc as red, green and blue components using 5 bits for
each color plus one overlay or control bit.

169

170 CD-I: A Designer's Overview

directly addressable sector A sector that may be addressed directly in
terms of time and number of sectors from a given reference point -
normally the start of the first data track on the disc.

directory record A record describing one file in a directory.

directory search method The method used to locate a specific file on
the disc. This method allows any file to be opened using only one seek.

disc bootstrap routine Optional routine on a CD-I disc to add or replace
operating system capabilities in a Base Case system.

disc interchangeability The ability to exchange discs between players
of different manufacture. This is an essential feature of both CD-DA and
CD-I.

disc file protection mechanism A method of allowing multiple
products to reside on a single CD-I disc and forcing the consumer to
purchase each product before it can be used.

disc label The information in the first track of a disc concerning the
disc type and format, the status of the disc as a single entity or part of an
album, the data size and the position of the file directory and boot
modules.

disc map The organization of data on magnetic tape as it will be on the
CD disc.

disc replication The production of copy discs from a master disc, usually
for commercial distribution.

disc storage Data storage on optical or magnetic disc, characterised by
low cost and relatively fast data access, compared with tape storage.

disc/system validation The procedure for checking the correct
operation of a CD-I disc or a CD-I system using a ’reference’ CD- I
system or ’reference’ CD-I disc respectively.

display (1) Generally, a device for presentation of visual information
which varies with time. (2) The visual image on the screen of a display
unit or monitor.

display controller Two-path device which takes pixel data from the
two banks of RAM of a CD-I player and combines them to produce a
single analog RGB video output.

Appendix B: Glossary of terms

display control program (DCP) A set of command codes are
interpreted by the display hardware during either the horizontal or
vertical retrace periods. The codes can be used to perform a variety of
functions; background colors, entries for color look- up tables, display
parameters, etc.

display resolution The measure of the number of pixels, and thus the
amount of detail, that a screen can display.

display unit Synonymous with video display.

dissolve The simultaneous fade-up of one image and fade-down of a
second image.

double-frequency scanning A method of scanning at twice the normal
frequency so that double the number of lines can be shown within one
frame, without the loss of quality or the line flicker of normal interlaced
scanning. Improves the vertical resolution of the display.

double resolution Twice as many pixels as a normal resolution image
in the horizontal direction, but the same number in the vertical direction.

double word A numeric entity comprising twice the number of bits
contained in a normal computer word. For a 16-bit processor the double

word is 32 bits wide; for a 32-bit processor it is 64 bits wide. Occupies
two successive memory locations.

DPCM See delta modulation.

drawmap A block of memory allocated by the User Communications
Manager to store image data.

driver Distinguishes hardware-specific features of CD-I players and
implements functions of file managers and isolates software for hardware
functions.

dynamic loading In CD-I, updating the contents of the color look-up
table (CLUT) during the horizontal retrace period (up to 4 colors) or
during the vertical retrace period (up to 256 colors).

DYUV See delta YUV

ECC See error correction code.

EDC See error detection code.

171

172 CD-l: A Designer's Overview

EFM See eight-to-fourteen modulation.

eight-to-fourteen modulation In Compact Disc, the pulse code
modulated signal produced by analog-to-digital conversion is a simple
non-return-to-zero bit stream of 1s and Os. It is not self-clocking, and
there is no restriction on Run-Length. (The number of successive 1s and
Os). To record this signal directly on to the disc would not only be
inefficient in terms of disc storage capacity. It would also make playback
very difficult, if not impossible.

Eight-to-Fourteen Modulation (EFM) is therefore applied, to produce a
signal format suitable for recording. EFM imposes a minimum
Run-Length of 3 bits and a maximum Run-Length of 11 bits. It also
changes the signal into a non-return-to-zero inverted bit stream, in which
a 1 is represented by a transition, and a 0 by no transition. Finally, EFM
introduces a unique synchronization pattern to each frame of audio
information.

EFM greatly reduces the number of transitions for the same amount of
data. This means that the data can be read more reliably, with much less
risk of interference between symbols. It also means that 25% more data
can be recorded on the same disc area. At the same time, EFM ensures
that there are always enough transitions to allow bit clock regeneration
in the Compact Disc player. The data is thus made self-clocking.

EFM also minimizes the difference between the number of 1s and the
number of Os in the bit stream. This suppresses low frequency
components which could otherwise interfere with the player’s focussing,
tracking and motor control servos.

Finally, the synchronization pattern allows each frame to be recognized.
This is essential, particularly for error correction and subcode separation.
EFM changes each 8-bit symbol in the signal into a 14-bit symbol. The
14-bit symbols all have a minimum of 3 anda maximum of 11 successive
Os. 256 such symbols are needed to match all the possible 8-bit
combinations. (In fact 267 14-bit symbols meet this requirement; 11 are
not used). The 256 14-bit symbols form a look-up table held ina RAM.
The Run-Length conditions must be maintained between symbols as well
as within them. This is achieved by inserting two merging bits. A third
merging bit maintains the balance between the number of 1s and Os in
the bit stream.

Thus, each 8-bit symbol becomes a 17-bit symbol (14 +3). The
synchronization pattern consists of 24 bits, and is uniquely identifiable.
It, too, has 3 merging bits.

An eight-to-fourteen modulated audio frame is composed of 33
seventeen-bit symbols (24 audio, 8 parity and 1 subcode) plus a 27-bit
synchronization pattern; a total of 588 ‘channel bits’. This is the signal

Appendix B: Glossary of terms

written on to the disc, where each 1 is represented by the beginning or
end of a pit. See also Compact Disc-Digital Audio, pulse code
modulation.

electronic publishing A generic term for the distribution of infor-
mation on computer databases linked by communication networks.
Videotex is an example of this method of publishing.

electronic storyboard An electronic version of a storyboard, usually in
the form of computer software program, used in the development of
among other things, a CD-I disc design.

end-of-file bit A bit in the submode byte of the subheader that is set to
1 for the last sector of a real-time file

end-of-record bit A bit in the submode byte of the subheader that is set
to 1 for the last sector of a real-time record.

EOF-bit See end-of-file bit
EOR-bit See end-of-record bit

erasable optical disc Research into erasable media suitable for use in
optical discs has been under way for some time at Philips Research
Laboratories and elsewhere. Two main technologies have so far
emerged. The first of these uses a technique known as magneto-optical
recording, and erasable optical discs have been produced experimentally.
Writing and reading depend on the physical effects of small,
reverse-polarised magnetic domains in a thin polarized magnetic layer.
Writing is performed by reversing the polarization of the domain, while
under the influence of an external magnetic field, by heating it above the
compensation point temperature with a short laser pulse. Reading is
performed by measuring the Kerr effect, which rotates polarized light
when it is reflected under the influence of a magnetic field.

The second method under investigation uses the inherant difference in
reflectivity of the crystalline form and non- crystalline (amorphous) form
of the same material as a starting point. The information is recorded by
rapidly heating small areas in a thin layer of crystalline material to
slightly above its melting point with a fairly powerful laser beam. These
small areas then solidify (the ’supercooled phase’). This produces
amorphous areas in a crystalline material and these can be detected
optically by the variation in reflectance. The differences in reflectivity
are quite sufficient for digital readout and sufficiently wéll defined for
the reproduction of analog video signals.

173

174 CD-I: A Designer's Overview

Because the crystalline form of materials is the most stable, all materials
tend to change into this phase. This effect can be used to erase
information on the disc. Heating to just below the melting point with a
laser beam will return the material to its fully crystalline state.

As a result of research work into amorphous-crystallinematerials,
gallium antimonide and indium antimonide were discovered as suitable
materials during the course of 1986.

These materials have a long shelf life, and are insensitive to ordinary
ambient temperatures and to humidity. Information can be erased and
re-recorded about a thousand times, which is quite sufficient for
consumer applications.

Research into these possibilities for erasable optical recording continues.
Improvements in the signal-to-noise ratio can be expected, as well as the
number of times that information can be erased and rerecorded.

error concealment A variety of techniques used in concealing errors in
visual images displayed from a CD-I disc.

error correction Identification and correction of errors arising in the
transfer of information. Used extensively in computer storage media
such as Compact Discs. See Cross-Interleaved Reed-Solomon Code,
cyclic redundancy check, error correction code, error detection code.

error correction code (1) In computing and communications, a code
designed to detect an error, in a word or character, identify the incorrect
bit and replace it with the correct one. (2) An error correction code used
in CD-ROM and CD-I to achieve high data integrity. See form 1, mode
1.

error detection code (1) In computing and communications, a code
designed to detect, but not correct, an error in a word or character. (2)
An error detection code used in CD-ROM and CD-I to achieve high data
integrity. See form 1, mode 1.

executable code A set of instructions, or a computer program, in the
machine language for a specific computer or microprocesor, and which
can be executed, or run, directly.

executable object code The output from a compiler’s or assembler’s
linkage editor or linker, which is in the machine code for a particular
processor, with each loadable program being one named file (module).
In CD-I, such an object file does not contain audio or video data. See
executable code.

Appendix B: Glossary of terms

extended disc In CD-I, a hypothetical disc that can exercise all the:

capabilities of an ‘extended’ system as defined by the CD-I extended’
system specification. The Base Case specification is a subset of the
‘extended’ system specification.

extended system In CD-I, a system conforming to Base Case
specification, plus any extensions that conform to the ’extended’ CD-I
system specification.

extension (1) In CD-I, an upward compatible module to replace an
existing system module in ROM. During initialization, all modules in
CD-RTOS (except the protection modules) may be replaced by extended
modules which have revision numbers higher than the ones they replace.
(2) In CD-I, a hardware module supporting a functional extension
conforming to the CD-I ’extended’ system specification. During
initialization, CD-RTOS identifies the extension and includes the
software modules from it.

fade A gradual decrease (fade-out) or increase (fade-in) of the brightness
of an image, or the volume of an audio signal.

field control table A one-dimensional array of instructions which are
carried out before the start of each field.

file A collection of logically related records identifyable in a directory.

file descriptor record A sector found in all CD-I files, containing a list
of the data segments, their starting logical sector number, size and file
attributes. Needed to access files on a disc.

file manager In CD-I, a system software module which handles I/O
requests for the CD drive. Provides random access to files at the byte
level through system calls.

file structure volume descriptor Record of the Disc Label describing
all necessary items related to the files or parts of the files recorded on the
volume.

file protection A mechanism whereby access to information in a given
file is resticted to those possessing the valid access key.

flowchart A chart indicating the interactive and pseudo-linear structure
of the disc.

175

176 CD-I: A Designer's Overview

form bit Bit in the submode byte of the subheader field defining the
data form (Form 1 or Form 2) for the sector.

form 1 The CD-I sector format with EDC/ECC error detection and
correction. Equivalent to CD-ROM mode 1, but with the form identity
included in a sub-header to permit interleaving of form 1 and form 2
sectors to meet the requirements of real-time operation.

form 2 The CD-I sector format with an auxiliary data field instead of
EDC/ECC error detection and correction. Equivalent to CD-ROM mode
2, but with the form identity included in a sub-header to permit
interleaving of form 1 and form 2 sectors to meet the requirements of
real-time operation.

frame (1) In computing the array of bits across the width of magnetic
or paper tape. (2) An individual picture on a film, filmstrip or videotape.
(3) A single television tube picture scan combining interlaced
information. (4) In videotex, a page of data displayed on a terminal. (5)
In CD-DA, one complete pattern of digital audio information,
comprising 6 PCM stereo samples, with CIRC and one subcode symbol,
eight-to-fourteen modulated, with a synchronization pattern.

frame grabber (1) In recording, an electronic technique for storing and
regenerating a video frame from a helical video tape signal. This method
avoids the need for the continuous -head to tape contact that would
otherwise be required in freeze frame operation. (2) An electronic device
for extracting a complete frame from a video signal and storing it in
memory for further processing.

full-motion video Not a design option in full screen for CD-I at this
time. However, a combination of the video bit set to 1, form bit set to 0
and real-time sector bit set to 1 is reserved for full-screen full-motion
video.

genlock A capability planned for CDI-X in which the background plane
will be synchronized to an external video source, allowing the CD-I
application to interact with external video information.

glass master An optical master disc produced by exposing a
photosensitive coating on a glass substrate to a laser beam, then
developing the exposed coating and covering it with a silver coating. See
CD-disc master, CD mastering.

global dissolve A dissolve affecting the whole of a video picture.
Compare local dissolve.

Appendix B: Glossary of terms

global fade A fade affecting the whole of a video picture. Compare local
fade.

global search A search operation performed on a complete file or
database. Compare local search.

glyph The bitmap image for a symbol. For example, the bitmap image
of the letter "a" is the glyph for the letter "a".

graceful degradation In CD-I, degradation of audio or video quality
due to increasing error content.

granulation An effect whereby the resolution of an image changes
without its size altering. It is produced by a combination of pixel hold
and line hold functions.

graphics cursor functions User Communications Manager functions
which control the shape, size, color and position of the graphics cursor
on the display screen.

graphics drawing functions User Communications Manager func-
tions which are used to draw images in drawmaps.

group execute (file attribute bit) This bit of the attribute field of a
directory record specifies that only users belonging to an identified group
can execute the program contained in this file.

group read (file attribute bit) This bit of the attribute field of a
directory record specifies that only users belonging to an identified group
can access this file.

green book Informal name for the CD-I specification.

hardware-dependent Of a system, dependent for its operation on a
specific hardware configuration.

header field In CD-ROM or CD-I, the second level of audio quality. A
bandwidth of 17000 Hz is obtained using 8-bit ADPCM at a sampling
frequency of 37.8 MHz. Comparable with LP record sound quality. See
audio quality level.

high resolution (1) The degree of detailed visual definition (800 x 600
pixels) that gives readable 80 column text display. (2) In CD-I, a display
resolution mode of 768 pixels (horizontal) by 560 pixels (vertical).
Compare normal resolution, double resolution.

177

178 CD-l: A Designer's Overview

High Sierra group An ad-hoc standards group set up to recommend
compatible standards for CD-ROM. The group includes representatives
from the hardware, software and publishing industries, and was named
after the hotel in Lake Tahoe where it first met in the summer of 1985.

horizontal line update The modification of all or part of a single line
in a video image.

horizontal retrace period Time during which the horizontal line scan
on a TV screen returns to the beginning of the next line.

hot spots Synonymous with active areas.

hyper-media An extension of hyper-text incorporating cross-linked
databases contained not only data in the form of text, but also in the form
of music, voice, sound effects, graphics, as well as still and moving
visuals.

hyper-text A series of logically interlinked databases, where the
information within one database can be logically cross-linked with the
related information in another.

icon Pictorial representation of an object in a graphic display.

idea map An overview of a CD-I disc project in the form of a map-like
structure, charting the relationships between individual elements making
up the disc content, in such a way that the links between previously
undefined elements can be established with a degree of confidence.

ideogram Symbol or character conveying an idea, expression or part
thereof (e.g. a Kanji character).

image A full or reduced screen image, or a partial update or an irregular
partial update.

image plane A displayed image may be formed by the superimposition
of a number of component pictures. Each of these constitutes an image
The CD-I system has a maximum of four image planes including the
cursor and backdrop planes.

information area Synonymous with recorded area.
information exchange protocol (IXP) A protocol for labelling and

describing data on a CD-I disc which is used in the mastering process.
This data describes the media type and encoding techniques, among other

Appendix B: Glossary of terms

characteristics. This data may also include other descriptions of the disc
data which might be useful to the mastering facility.

information carrier Any medium by which information is carried from
its point of origin to its point of use, e.g. magnetic tape,Compact Disc,
transmission line, broadcasting channel, or paper.

input port An interface used for transferring information into a
computer. See I/O port.

insert module A module which, when inserted into an equipment or
system, enables it to perform additional functions.

intelligent player A CD player with additional computing facilities built
in, enabling the player to interact with the user, or to operate under
program control. CD-I players are intelligent players.

interactive LaserVision A system which employs a LaserVision drive
and a (micro)computer, either built-in or external, to run interactive
programs from a CAV type LaserVision disc.

interactive medium Medium which presents information in such a way
that, by means of an application program, it is delivered in the course of
a dialog with the user. The application program may also be included in
the medium. Examples include Interactive LaserVision and CD-I.

interactive mode Presentation of information in a sequence determined
by a dialog between the information medium and the recipient.
Compare linear mode. See interactive medium.

interactive system A system capable of using an interactive medium to
supply information to the user.

interactive video (IV) Synonymous with interactive LaserVision.

interlace A system of picture scanning using two fields, the lines of the
second field being interposed between those of the first. Interlace
scanning produces a higher level of detail while minimizing the flicker
inherent with low refresh rates.

interleaving In CD-I, the process of physically separating data so it can
be retrieved at the rate required for processing the data. It involves the
interspacing of sectors at intervals that correspond to the nature of the
data. For audio, a regular interspaced pattern is used which depends on
the audio quality level required. The subheader indicates the interleaving

179

180 CD-I: A Designer's Overview

pattern at file, channel and data type levels. Blocks between consecutive
audio blocks can be used for video or text data.

International Standard Recording Code A code used by record
manufacturers. Gives information about country of origin, owner, year
of issue and serial number of individual music tracks. May optionally
appear in CD-DA subcode.

interpolation scheme A method of concealing errors.

InVision A user interface toolkit developed by Microware. Most CD-I
players will have InVision in ROM.

I/O functions In CD-I, the transfer functions Read and Play which
perform the physical transfer of data from the disc.

1/O port An opening on a CD-I player or microcomputer enabling an
external device to be connected for input/output operation.

irregular updates One of the two types of partial updates available
within CD-I, where the area to be updated is irregular in shape.

ISRC See International Standard Recording Code.

ISO 646 International Standard specifying the 7-bit coded character set
for information interchange.

ISO 2022 International Standard specifying the methods (through shift
function or escape sequences) to extend 7 or 8 bit coded character sets.

ISO 2375 International Standard specifying the registration procedure
of escape sequences used for code extension of a character set. It refers
particularly to the International Register of coded character sets to be
used with escape sequences.

ISO 8859/1 The specification of the coded character set for Latin
alphabet No. 1 used as the standard character set definition for CD-I
machines.

IV Interactive Video. Synonymous with interactive Laser Vision.

IXP Information exchange protocol.

kernel The nucleus of CD-RTOS.

Appendix B: Glossary of terms

keyboard input functions User Communications manager functions
which are used to get character data from a keyboard device.

LaserVision Optical video disc system developed by Philips for
reproducing and 2-channel sound.

lead-in area In CD, a track (number 0) on the disc preceding the program
areas. Contains the table of contents.

lead-out area In CD, a track (number $AA) on the disc following the
program area.

line control table A two-dimensional array of instructions, each row of
which is associated with the displayed line.

line hold, line repeat In the vertical direction, mosaic graphics are
produced by holding and repeating lines.

line multiplication A technique used in CD-I to make high-resolution
line information compatible with a lower-resolution system.

linear mode Presentation of information in a fixed sequence, un-
influenced by the recipient. Compare interactive mode.

line update The modification of single line, or part of a line, of graphics
stored on a file.

linker See linking loader.
load time In video, the time taken to put acomplete picture on the screen.

loadable program A single named file containing object code
information used in CD-I.

local dissolve A dissolve affecting a portion of the image. Compare
global dissolve.

local fade A fade affecting a portion of an image. Compare global fade.

local search A search operation confined to part of a file or database.
Compare global search.

low resolution A degree of detailed visual definition below the 400 x
300 pixels presented by normal domestic color TV sets. Compare normal
resolution, double resolution, high resolution.

181

182 CD-I: A Designer's Overview

luminance The measured radiance of a light source expressed by Y
signals. Compare chrominance.

LV See LaserVision.

master disc An original disc, from which copies can be made by a
replication process. See CD Disc master.

mastering In optical disc, the production of the master disc.

main channel address The address in time, seconds and sector numbers
on the main channel of a given sector.

matte An area on an image plane which is made transparent, permitting
the image behind it to show through. A matte can also be defined so that
the area itself remains visible, while everything around becomes
transparent.

media palette In CD-I, the available range of technology functions for
storing information on a CD-I disc, for use in a range of specific design
applications.

memory module A named block of executable program statements or
data that can be loaded by CD-RTOS into memory.

metal father A recording mould formed by nickel plating on a master
disc. Can be used directly for replication, or as the basis for the
production, by two further stages of plating, of stampers for large-
quantity production.

menu-driven The course of events in an application program, inter-
actively controlled by means of menu selections.

message sector Sector containing caution encoded as CD-DA audio
data to warn the user of a CD-DA player to lower the volume.

mid fi quality In CD-I, the third level of audio quality. A bandwidth of
12000 Hz is obtained by using 4-bit ADPCM at a sampling rate of 37.7
kHz. Comparable with FM broadcast sound quality. See audio quality
level.

mix The combination of two (or more) images into a single image.

mode 1 One of the two physical sector formats defined for CD-ROM.
Incorporates EDC/ECC error detection and correction.

Appendix B: Glossary of terms

mode 2 One of the two physical formats defined for CD-ROM.
Incorporates an auxiliary data field instead of EDC/ECC error detection
and correction.

mode byte In CD-ROM, the byte in the header field of a sector that
defines whether a sector is mode 1, mode 2 or mode FF.

mode FF [XP information

mosaic graphics In CD-I, low-resolution graphics achieved by repeating
pixels or lines by a certain factor.

mother In disc replication, anegative mould intermediate between metal
father and stamper. Formed by nickel plating on the metal father. See
metal father.

multiplane A video image in which various different pictures are
overlaid one on top of the other.

natural images Pictures which are photographic in nature and appear
realistic.

natural pictures See natural images.

near-instantaneous compression Compression performed quickly
enough to have no perceptible effect on the timing of the presentation of
the information concerned.

new media Media now becoming available, or envisaged as becoming
available, for mass information presentation.

non-linear quantization Quantization using steps of different sizes, to
distribute the steps more efficiently over the dynamic range. Takes
advantage of the fact that quantization errors are less perceptible when
signal changes are large.

non volatile random access memory Memory able to retain its contents
when the main power to the unit is removed.

normal resolution In CD-I, a display resolution mode of 384 horizontal
pixels by 280 vertical pixels. The lowest resolution picture defined in
the CD-I system. Compare low resolution, high resolution, double
resolution.

NV-RAM Non volatile random access memory.

183

184 CD-I: A Designer's Overview

omni player A CD-Video player that will play CD-I discs.

object code The code of a user’s program after it has been translated by
means of an assembler or a compiler.

odd/even line separation When error concealment is to be used, the
even and odd lines of an image are coded on a disc in separate sectors
and are physically separated by interleaving so as to minimize the chance
of an error occuring on adjacent lines.

on-board Additional or supporting function incorporated into a
printed-circuit board or within the housing of an equipment.

one dimensional Run-Length coding A picture coding technique in
which pixel data is compressed using Run-Length coding in the
horizontal direction only.

optical digital disc An optical disc in which information is stored
digitally.

optical disc A disc in which information is impressed as a series of pits
in a flat surface, and is read out by optical means, i.e. by a laser.

optical input In optical media, the light signal before it is converted into
an electrical signal.

optical medium Medium employing optics for the storage and
distribution of information.

optical recording The recording of information in such a way that it can
be read by a beam of light. Modern optical recording technology is almost
entirely concentrated on the use of low-power lasers to write and read
information on optical discs. Optical discs can carry very large amounts
of information per unit volume. They are highly resistant to damage and
immune to electromagnetic influences. Access to information is fast and
error rates can be made very low. Laser read-out completely eliminates
wear during use.

DOR Digital Optical Recording systems can write and read digital data
(though at present without any erasure facility). Their principal
application is in large-scale archiving.

Rapid developments are being made in read-only optical recording
systems, which include CD-DA, CD-ROM and CD-I.

Appendix B: Glossary of terms

optical storage Storage of information in such a way that it can be read
using optics. Characterised by very high storage density.

optical technology Technology based on the use of optical effects for
the transmission or storage of information.

OS-9 The real-time operating system which forms the basis for the CD-I
operating system CD-RTOS.

output port An interface used for transferring information out of a
computer. See I/O port.

overlay The process of superimposing image planes in a given visual
image.

overlay control In CD-I, the mechanism which controls transparency
between planes.

overscan Extending the deflection of the electron beam of a cathode ray
tube (CRT) is made-to extend beyond the usable physical dimensions of
the screen. In this way images will always fill the visible part of the
display screen.

owner execute (file attribute bit) This bit of the attribute field of a
directory record if set to one specifies that only the user identified as
“owner” of the file can execute the program contained in this file.

owner read (file attribute bit) This bit of the attribute field of a directory
record specifies that only the user identified as "owner" of the file can
access this file.

palette A range of colors. In CD-I, a palette is used by the User
Communications Manager to support the color look-up table. The
maximum size of the palette at any instant in time is 256 colors, with the
red, green and blue components each defined to 8-bit accuracy. See
media palette.

panning The distribution of a mono signal between left and right audio
channels.

parent directory A directory which contains sub-directories.

partial update New image data written to part of a picture that is
currently being displayed.

185

186 CD-I: A Designer's Overview

path descriptor A data structure used to represent an open file; each
open file is associated with a path descriptor.

path table Table or index used for directory search. Describes an entire
directory structure on a disc. Permits any file on the disc to be opened
using only one search.

P channel One of the eight Compact Disc subcode channels (P-W). The
P channel carries the music flag, indicating presence of absence of a
music track.

PCM See pulse code modulation.

PCM audio (1) A pulse-code-modulated audio signal. (2) In CD-DA the
audio signal after the first stage of encoding i.e. a multiplexed signal with
six 32-bit stereo samples in each audio frame.

photosensor Any device for converting light into an electrical signal.
picture element Synonymous with pixel.

Pixel The smallest element on an image which can be manipulated or
identified. Synonymous with picture element.

pixel aspect ratio The ratio of width to height of a displayed pixel.
Ideally this should be 1:1.

pixel hold Method used to reduce resolution. This function operates
after the pixel codes have been decoded to RGB.

pixel pairing Pixel pairing occurs in 3-bit run-coded images and 4-bit
CLUT encoded pictures. Two pixels are put together to make up one
byte, and are then regarded for most purposes (e.g. the length of the run
in run-coded pictures) as a single unit.

pixel repeat A visual effect function that repeats pixel values from an
image memory to produce horizontal magnification. Used in RGB 5:5:5
and CLUT image decoding.

plane See image plane.
play control block A data structure indicating what information is to be

accessed on the disc for the current or next real-time record play function
of the compact disc file manager.

Appendix B: Glossary of terms

play control list A structure list that controls the destination of the data.

players default character set Set of characters selected by the current
user of the player when a disc is initially accessed. For CD-I this
conforms to ISO/DIS 8859/1.

prediction filters The filters used in ADPCM encoding to achieve
effective response to audio frequency distribution fluctuations.

pre-mastering The stage between authoring and mastering.
program carrier Material or device to carry or store a program.

program area The area of a Compact Disc containing the program and
consisting of a maximum of 99 audio or data tracks.

program related data Data in the form of executable object code to be
read and processed by the CD-I MPU.

programmable sound generator Audio signal generator with an
integrated microprocessor to control the output signal according to a
program set up by the user.

pulse code modulation (PCM) A technique for converting analog
information into CD-DA digital form. The analog signal is sampled at
arate equal to at least twice the maximum signal frequency component,
and the sampled value is represented by afixed length binary number.
This number is then transmitted as a corresponding set of pulses.

Q channel One of the eight Compact Disc subcode channels (P-W). The
Q channel carries the main control and display information. It identifies
tracks, indexes and running times, and the absolute playing time of the
disc. It also indicates whether the recorded information is audio or data,
whether pre-emphasis is applied and whether digital copying is
permitted. It can also indicate 2 or 4 channel audio, should 4-channel
audio be introduced. Optionally, it can include a disc catalog number and
ISRC information. Finally, it includes its own CRC (cyclic redundancy
check).

QHY See quantized high Y.

QHY quantization levels The QHY difference values, which need to
be added to the normal resolution quantities to generate a pseudo-high
resolution image, are 8-bit quantities varying between 0 and 255. Eight
values are chosen from these 256 possible quantities. Each difference

187

188 CD-I: A Designer's Overview

value is then made equal to the nearest of these eight, and a three bit
number is used to represent it. The eight chosen values are known as
quantization levels.

quantize To assign one of a fixed set of values to an analog signal as
part of an analog to digital conversion process e.g. in pulse code
modulation, an analog signal is sampled, quantized and a corresponding
set of binary pulses is produced. See pulse code modulation.

quantized high Y A coding technique used to reduce the quantity of
data required to encode a high resolution type picture. A normal
resolution DYUV image is recorded together with the data which needs
to be added to it to turn the luminance (Y) of the picture to the equivalent
of high resolution. The latter data further compressed is termed ’QHY’.

quantizer A device or circuit which assigns fixed binary values to
sampled analog signals in analog-to-digital conversion.

quick disc Optical disc produced on very short delivery (normally
during one working day) to a special requirement or for program
validation prior to disc replication.

RAM Random access memory.

random access memory In computing, (1) a memory chip used with
microprocessors, information can be read from, and written into, the
memory but the contents are lost when the power supply is removed, (2)
any form of storage in which the access time for any item of data is
independent of the location of the data most recently obtained.

random area update An area of any shape, updated as a succession of
horizontal whole or partial line updates.

random error In computing, a spontaneous bit error, usually not re-
producible and independent of the data. This type of error is caused by
device operation up to physical boundaries.

raw sector A sector of information that includes the subheader, data and
error correction information.

read/write medium A medium that can be both written (record) and read
(playback). Magnetic media can generally be written, read, erased and
re-written repeatedly. Optical carriers are at present read-only, or write
once, read many times (WORM). Erasable optical discs are the subject
of intensive research.

Appendix B: Glossary of terms

read error An error in the reading of data from a storage medium.

real-time control area Disc sector preceding some real-time records
that contains all control data and specific instructions for playback.

real-time data In CD-I, data taken directly from the disc, whose flow
cannot be interrupted or stopped within the bounds of a real-time record.

real-time file A file containing at least one real-time record.

real-time interactive system An interactive system which responds to
events directly as they occur, i.e. in real-time. CD-I is a real-time
interactive system.

real-time operating system An operating system that functions within
the constraints of real-time, e.g. CD-RTOS, the OS-9-based operating
system of CD-I. Such an operating system is essential for full
interactivity.

real-time record In CD-I, the smallest amount of real-time data that can
be randomly accessed. A logical record in a CD-I file containing audio,
video, and/or computer data that must be retrieved from a CD-I disc ata
precise rate.

real-time record interpreter A trap-handler module which an
application can use to assist in the playback of real-time records.

real-time sector In CD-I, a sector with the real-time bit set. The data in
this sector must be processed without interrupting the real-time
behaviour of the CD-I system.

record The logical component(s) of a file.

recorded area The total area of the recording on a Compact Disc
including the lead-in area, the program area and the lead-out area.

rectangular update One of the two types of partial updates available
within CD-I, where the area to be updated is in rectangular form.

red book Informal name for the CD - DA specification.
reduced resolution The magnification effect produced by pixel and line

repeat may be used to produce a low resolution image allowing fast
update from the disc of a larger than normal screen area. This is an

189

190 CD-Il: A Designer's Overview

alternative to animation, like Run-Length compression, which may be
chosen for specific applications.

reduced screen display A display consisting of the safety area
surrounded by an appropriate border.

refresh rate The frequency with which new visual information can be
put up on a screen.

region A software mechanism in the User Communications Manager
which is used to limit the area of drawing in a drawmap, and to set up
mattes.

region generation In video, an overlay technique defining the overlay
area separate from the image contents.

replication The production of copies from a master, usually for
commercial distribution.

retrieval The process of searching for, locating and reading out data.

RGB encoding A video encoding technique which transforms the red,
blue and green components of a video signal into a PCM signal.

RGB (5:5:5) One of the video encoding techniques used for images in
CD-I. For each pixel the colors (red, green and blue) are each quantized
and represented by 5 bits of information, giving 32 levels of intensity
from one extreme value to the other. See also absolute RGB components.

ROM sce read-only memory

root directory The highest level directory contained on a disc. A root
directory must reside on every CD-I disc.

RL See Run-Length.
RLE See Run-Length (en)coding.

rotational latency The time taken for the disc to rotate under the read
head until the required block becomes available for reading.

RTCA See real-time control area.

RTOS See real-time operating system.

Appendix B: Glossary of terms

RTR See real-time record.
RTRI See real-time record interpreter.
Run-Length In a data stream, the number of bits between transitions.

Run-Length (en)coding A picture data compression technique which
uses two-byte codes. The first byte identifies color, and the second byte
tells the decoder how many pixels are to be of this color.

running time In CD-DA, the time that an audio track has been running.
Included in the subcode and thus available for display during playback.

safety area That part of an image (expressed in terms of pixel
coordinates) that is guaranteed to be displayed despite all tolerance
values that can occur in monitors and TV sets.

sample rate converter Unit which converts the sampling rate of a
digital audio signal.

SAT Stanford Achievement Test. Used by US Universities as a measure
of eligibility for University entrance

scan convertor A device used in presenting a non-interlaced picture on
anormal TV screen containing the same number of lines as the original
interlaced picture. See also double-frequency scanning.

scrambling In CD-I, all data except for the data in the synchronization
field is scrambled. The contents of a 15-bit shift register scrambler is
EXOR-ed with the serial information bit by bit.

scramble register A register used in the scrambling and descrambling
process for all data in a CD-I sector.

screen (1) Video display. (2) Displayed image

scroll The repeated repositioning of a displayed image within a large
image. Motion may be vertical, horizontal or a combination of the two.

search program A program that searches a data file or database to find
a keyword or key phrase supplied by an operator or another program.
See retrieval.

sector The smallest unit of absolutely addressable information in a
CD-ROM or CD-I disc. A sector is 2352 bytes long containing a

191

192 CD-l: A Designer's Overview

synchronization pattern, header field and digital data. It may also contain
a sub-header and EDC/ECC error protection. See sector structure.

sector address In CD-ROM and CD-I, the physical address of a sector
expressed in minutes, seconds and sector number. Contained in the
address part of the sector header.

sector structure In CD-ROM and CD-I, the 2352 sequential bytes of a
sector may be divided in one of four ways, depending on the system and
the degree of data integrity required. See the accompanying table.

CD-ROM and CD-I sector structure

CD-ROM CD-I

mode 1 mode2 form 1 form 2
Synchronization 12 B 12B 12B 12B
Header 4B 4B 4B 4B
Subheader - - 8B 8B
User Data 2048B 2336B 2048B 2324B
EDC/ECC 288 B - 280 B -

See form 1, form 2, mode 1, mode 2.

seek The action of changing the location of a pointer and then finding
the specified location.

seek latency Delay between a request for search action and arrival at the
location sought.

service request A request made of the kernel by an application to
perform a specific activity.

simulation A mock-up of how a CD-I application will play, using
whatever hardware and software may be necessary.

single-medium system In computers, a system architecture based on the
use of a single medium which carries all the software needed for a given
application. In CD-I for example, all the program data (video, sound,
text and computer), application and driver software is held on the CD-I
disc itself. Only the basic operating system kernel is stored - in ROM in
the base case CD-I player - external to the disc.

Appendix B: Glossary of terms

sleep A process execution state where the process will be inactive until
a specific amount of time has passed or an interrupt is received.

sound attribute In CD-I, a particular property assigned to all or part of
the sound information e.g. language, bandwidth.

sound group In CD-I, part of the user data field in an ADPCM audio
sector. Contains four sound units, each consisting of four identical sound
parameter bytes of 8-bits and 28 data bytes. Each ADPCM audio sector
has 18 sound groups.

sound macro A predefined sound or sound sequence stored in computer
form.

sounding Memory area that preloads audio data from disc.

soundmap A memory area allocated by UCM for storage of ADPCM
audio data.

sound parameters Filter and range values describing the character-
istics of a sound group.

sound quality level Four levels of audio quality are defined. These are
CD-Digital Audio with a bandwidth of 20 kHz and 16-bit sampling, CD-I
Level A, 17 kHz bandwidth with 8-bit sampling, Level B, 17 kHz
bandwidth with 4-bit sampling and Level C, 8.5 kHz bandwidth with
4-bit sampling.

sound unit A unit consisting of sound parameters and sound data bytes.

spatial correlation A characteristic of visual images that there is a high
degree of similarity between two adjacent images in a sequence of given
images.

speech quality In CD-I, the fourth level of audio quality. A bandwidth
of 8.5 kHz is obtained using 4-bit ADPCM at a sampling frequency of
18.9 kHz. Comparable with AM broadcast sound quality. See audio
quality level.

sprites Small images on a screen, movable under program control, and
normally ranging from character sets or cursor shapes to specific patterns
as in computer games.

stamper A recording mould used to press gramophone records or optical
discs.

193

194 CD-I: A Designer's Overview

storage capacity The amount of data a particular store can accommo-
date, generally specified in bytes. Storage can also be quantified in terms
of the type of information stored. The over 660 MByte storage capacity
of a Compact Disc, for example, can hold the data needed to reproduce
over 160,000 pages of typed text, 72 minutes of the finest quality sound,
or some 5000 video-quality natural pictures.

medium storage capacity access
capacity times
(bytes per
unit)
RAM 64k...83MB <1 sec
floppy disk 128kb...4MB 30msec
hard disk SMB...200MB 30msec
magnetic 2S5MB...1200MB minutes
tape
Compact Disc 650MB <1 sec

straight PCM mode In ADPCM, the PCM mode used as predictor for
signals having high frequency characteristics.

subcode channel In Compact Disc, one of eight sub-channels, referred
to as P to W, which exist in parallel to the main channel. They are used
for control and display information.

subheader In CD-I, a field indicating the nature of the data in a sector,
thus allowing real-time interactive operation. See sector structure,
synchronization. The field defines file number, channel number,
submode and coding information. The subheader can be thought of as a
series of real-time switches that reduce load on the microprocessor and
co-processors and save decoding time.

submode byte The submode byte defines global attributes of the sector.

subscreen The horizontal portion of a single frame that can use different
decoding and resolutions than other subscreens in the same plane.

superimpose Place a computer-generated image over an image from
another source. See overlay.

switching overlay A technique in which every pixel in the displayed
image is selected from one or other of the corresponding source images.

Appendix B: Glossary of terms

symbol In Compact Disc, the basic unit of digitized data, parity and
subcode data. Initially 8 bits long, is expanded to 17 (143) bits by
eight-to-fourteen modulation.

synchronization The process of maintaining common timing and
coordination between two or more operations, events or processes. In the
CD-I system, featuring simultaneous pictorial, sound and text
information, the synchronization of the various elements which form the
total presentation is an essential task of the applications program, under
the control of the CD-I operating system, CD-RTOS.

The data stream from the disc, which carries the information to be
interpreted by the CD-I player for presentation on the video screen and
reproduction by the hi fi system, consists of a series of sectors. A
subheader at the beginning of each sector, directly following the data
stream synchronisation and header fields indicates to the CD-I
controlling microprocessor the nature of the information in the user data
block which directly follows the subheader information. This user data
information can be part of the application program (or the boot or start-up
information for the application). It can be data for interpretation by the
video processor as pictorial information, or by the audio processor as
audio information. Or it can be text or other program data to be
interpreted by the main microprocessor.

Based on the indication contained in the sub-header, the microprocessor
switches the user data block to the appropriate circuit.

It is then the task of the application program to instruct the micro-
processor how to handle the information once it has passed through the
relevant decoding process. In some cases, such as for CD-DA music
tracks, the output data will be switched directly to the audio output
channels.

In the case of applications program or computing data, the information
may well be stored in the main memory, while video data will pass to
the video memory to build up a picture for later display.

The synchronization function of the application then relates the various
outputs from these data buffers to data coming directly from the
appropriate decoding circuitry, to ensure that they are all presented in
correct synchronization.

In the example illustrated, the synchronization data control triggers the
video memory to indicate when the picture transfer from the disc is
complete, and then to pass the background picture of the clouds, house
and earth to the screen. At the same time, the sound data representing
rain is passed to audio channels. At the appropriate moment, the overlay
of the lightning flash is triggered to the video output, and a short time
later, the related sound of thunder is passed to the sound channels.

A second example shows a cooking program, with the additional
complexity of multilanguage text in synchronization with multilanguage

195

196 CD-I: A Designer's Overview

speech. The other elements which make up the presentation -
background, clock, smoke and a moving head and body - are all
synchronized in a similar manner to the first example.

synchronization field In CD-ROM and CD-I, the first 12 bytes of a
sector containing synchronization information.

synchronization primitives Low level mechanisms for use by
application software to perform synchronization of the video, audio and
data sectors with themselves.

synchronization signals In CD-I, real-time software interrupts, often
generated by a device driver during hardware interrupt processing when
a predefined condition has been met.

synthesis parameters The parameters used to regenerate audio
information from data stored in a compressed or encoded format.

system modules Program modules that are required for the CD-I
system to function. These include the kernel, file managers and device
drivers.

system state A special state of the processor which allows execution
of privileged instructions; synonymous with the "supervisor" state of the
68000 family of microprocessors. File Managers and device drivers
always execute in system state.

system text In CD-ROM and CD-I, amessage processed by the operating
system without the need to load and process any special text processing
application program.

table ofcontents Subcode information defining the sequential number,
start, length and end times of tracks on a Compact Disc, together with
their type, i.e. digital audio or data. The table of contents is contained in
the Q sub-code channel of the lead-in area of all Compact Discs.

TOC See table of contents.

track (1)In recording and computing, a path along which data is
recorded, on a continuous or rotational medium, e.g. magnetic tape,
magnetic disc. In video recording the track is diagonal on the tape. In
magnetic discs the data is recorded on a series of circular tracks. (2) In
Compact Disc, a sequence of contiguous data, the beginning, length,
mode and end of which are defined in the table of contents, which is held
in the Q subcode channel of the lead-in area of the disc. The two types

Appendix B: Glossary of terms

of tracks currently defined are the CD-DA track according to the CD-DA
specification and the data track according to the CD-ROM specification,
which is also used in CD-I. In CD-DA the length of a track is related to
playing times between 4 seconds and 74 minutes.

tracking The following of a track by a readout or pick-up device.

transition (1) In filming and video, change from one image to another.
(2) In digital technology, change of state in a bit stream.

transition area A short sequence of DYUV codes which have been
pre-calculated to make the transition from the YUV values in the
surrounding image just preceding a partial update to those at the left edge
of the update proper. Each line of a partial update must similarly be
terminated with a transition area.

transparency bit In CD-I, a dedicated bit controlling overlay
transparency in the cursor plane and the RGB (5:5:5) plane.

transparency control The three transparency mechanisms used to
control the display of superimposed image planes.

treatment An overview of a proposed CD-I title, including information
on the logistics involved in the title’s creation.

trigger, trigger-bit A bit in the submode byte of the subheader that is
interpreted by an application to cause synchronization of various events.

U One of the two chrominance components of a video signal containing
color information.

UCM See User Communications Manager

user application A program or related set of programs designed for the
user of a system, rather than for programmers or service technicians. See
applications software.

user communications manager The CD-RTOS file manager which
provides the software interface to the user and output devices on a CD-I
player.

user data In CD-ROM and CD-I, data supplied by an information
provider for an application. As such, includes retrieval software, but not
information the information provider may be required to supply to
facilitate authoring.

197

198 CD-l: A Designer's Overview

user data field In CD-ROM and CD-I, a 2048-byte-long portion of the
data field in an addressable sector, dedicated to user data.

user interface The interface through which the user and a system or
computer communicate. Includes input and output devices such as a
keyboard, a hand control, a touch pad, a touch screen, aprinter and a
display, and also the software-controlled means by which the user is
prompted to supply data needed by the application, and by which he is
notified of his errors and how to correct them.

user number A code or password by means of which an authorised user
can gain access to acomputer or to stored information. For example, the
PIN, or Personal Identification Number used for authorising bank
transactions.

user shell In computers, a program between the operating system and
application program on the one hand, and the user on the other, to
enhance the manner of information presentation and command.

user state The state of the machine when user processes execute and
user related requests can be processed.

V One of the two chrominance components of a video signal containing
color information.

vertical retrace period Time during which the vertical field scan on a
TV screen returns to the beginning of the next field.

video data In CD-I, data related to one or more units of video information
as encoded in DYUV, RGB (5:5:5), CLUT or Run-Length encoding
techniques.

video data sequence The basic unit of image data loaded by a command
in the real-time control area (RTCA). It typically contains the pixel data
for an image or a partial update to an image.

video error concealment In digital video, a technique to reduce the
visual effect of disturbances arising from erroneous video data.

video input-output The facility for video input as well as output from a
computer. With frame grabbing, for example, video signals can be input
to the computer for additional processing, and then output to the display.
See frame grabber.

Appendix B: Glossary of terms

video quality level The reproduction quality of a video signal. CD-I
provides for four video quality levels viz. natural pictures, RGB (5:5:5)
graphics, CLUT graphics and Run-Length-coded animation.

visual effects function In CD-I, one of the set of functions, such as signal
mixing and color palette control, used to achieve visual effects.

voice grade audio information Audio information ofa quality sufficient
for reproducing the human voice, normally having a bandwidth of 4-8
kilohertz. See speech quality.

volume A disc that forms part of a set, or album.

volume identifier Field of the File Structure Volume Descriptor
identifying the name of the CD-I disc (volume)

wipe The replacement of one image by another during a period of time
by the motion of a boundary separating the visual parts of the two images.

world disc A CD-I disc on which the video data is encoded in such a
way that it can be played and displayed on any CD-I player, irrespective
of 525 or 625 line TV standard.

world execute (file attribute bit) This bit of the attribute field of a
directory record if set to one specifies that any user can execute this file.

world read (file attribute bit) This bit of the attribute field of a
directory record if set to one specifies that any user can access this file.

WORM See write-once, read-many(-times) medium.

write-once medium Medium on which data, once written, cannot be
erased to permit re-writing e.g. card, paper tape and DOR optical disc.

write-once, read-many (times) medium Synonymous with write-once
medium.

write/read medium See read/write medium.

X-Y device Input device for entering X and Y coordinates, mainly used
for accurate cursor positioning.

Y component The luminance or brightness component of a video
signal.

199

200 CD-I: A Designer's Overview

yellow book Informal name for the CD-ROM specification.

YUV In video, symbol denoting the luminance signal (Y) and the two
chrominance signals (U and V). See YUV encoding.

YUV encoding A video encoding scheme taking advantage of the human
eye’s reduced sensitivity to color variations as opposed to intensity
variations. In each picture line, the luminance (Y) information is encoded
at full bandwidth, while on alternative lines the chrominance (U and V)
signals are encoded at half bandwidth.

zoom In video and photography the facility to enlarge, (zoom-in) or
diminish (zoom-out) the area of interest in an image.

Appendix C: Introduction to CD-RTOS and Invision 201

APPENDIX C: INTRODUCTION TO CD-RTOS
AND INVISION

CD-RTOS is the operating system component of the CD-I system. It
stands between application programs and the details of the hardware. If
there are several programs running at once, it insulates them from one
another.

CD-I is new, but CD-RTOS is a new name for a familiar and thoroughly
tested operating system, OS-9. It was written for the 6809
microprocessor in the late 70’s, and ported to the 68000 in 1983. Both
the 6809 and the 68000 versions of OS-9 are widely used in process
control and personal computing applications. Versions of OS-9 are
available for several popular personal computers including the Tandy
Color Computer and the Atari ST. The name is new and CD-RTOS
includes extensive new I/O support, but the the operating system has
years of experience behind it.

Most of the large body of OS-9 software can be adapted to CD-RTOS.
These programs should run under CD-RTOS with no more changes than
might be required to adjust them to a new terminal type. From the
operating system’s point of view the sophisticated CD-I hardware is
"only" a collection of new I/O devices.

The diverse history of CD-RTOS brings some unusual functions to the
CD-I player. The CD-I base case system will contain an operating system
with the latent ability to turn the system into a serious computer. It seems
improbable to use a CD-I player as a multi-user computer, but all that is
lacking is a writable mass storage device (e.g., a floppy disc drive), and
plugs for a few terminals.

The CD-I system does not simply copy data from the disc to its output
devices. Except when it is emulating a CD-audio player, a CD-I system
expects the application to supply a program that will run under
CD-RTOS.

The program controls and modifies the data flowing through the system.
It is the last point at which a CD-I designer can have control of the data,
and the only point where the actual sequence of events in an interactive
application is known. Imagination and craftsmanship in programming
will improve the responsiveness of the application.

202 CD-|l: A Designer's Overview

BASIC CONCEPTS

Processes

Operating systems like MSDos and the Macintosh operating system
don’t support multitasking. They can be forced to run a print spooling
program, but only by following special rules while writing the spooler
program. CD-RTOS is designed to run a large number of programs at the
same time. There are no special rules or restrictions on print spoolers, or
compilers, or clock displays, or any other programs that you might want
to run concurrently.

The CD-I hardware requires multitasking support. There are four
separate streams of data coming off the disc, plus akeyboard and pointing
device that may be used at any time. While the system takes input from
these six sources it must be able to simultaneously deliver output to the
video and audio processors. All eight I/O activities must take place on
demand. The CD-RTOS operating system handles much of the problem
of keeping eight I/O balls in the air. Servicing all the streams of input
and output seldom uses all of the processor’s power, and the processor
time that is not used by CD-RTOS is available to application programs.
These programs need not use more than one task (tasks are called
processes under CD- RTOS), but multitasking is the best way to ensure
that all available processor time is used.

A skillful programmer can use multitasking and the other asynchronous
CD-RTOS services to make applications run more quickly. If a program
needs to wait for keyboard input, the programmer should ask himself
whether there is some way to make use of the processor time until the
input arives. Under CD-RTOS waiting for input doesn’t use any
processor time and even a fast typist doesn’t enter more than about ten
characters a second. The CD-I machine can do a great deal of work ina
tenth of a second.

Modules

Programs are made up of machine instructions - code - and data. Code
is kept in objects called modules which can be stored on discs, in ROM
(Read Only Memory), or in RAM (Random Access Memory). Every
module has a name, and when the module is in ROM or RAM the
operating system can use a module’s name to locate it.

Programs consist of one or more modules. The modules making up a
program may be shared among several programs. They can also be
loaded into memory when needed and deleted when a program needs the
memory for something else. Both of these tricks save memory.

Appendix C: Introduction to CD-RTOS and Invision

Programs written for CD-I systems should be exceptionally flexible. If
the program is written as a number of modules it can be flexible without
wasting memory on all possible options. For instance, a program could
choose a display support module depending on the user’s preferences
(language, type of display, type of pointing device, and whatever else).
The programmer would have to write separate modules for each display
mode the designer wanted to support, but all those options would need
to be programmed for in any case. By loading a tailored module the
programmer can save the memory which would have been consumed by
all the other alternatives. The program will also run faster by saving the
time it would have spent selecting the right alternative every time display
services were used.

The CD-RTOS operating system is built of a number of modules. This
gives it tremendous flexibility. Adding support for new hardware to the
operating system is a matter of changing or adding a few modules. If a
vendor wants to sell a hard disc drive for a system running CD-RTOS,
they will need to include two or three modules with the drive. The
customer can connect the drive, load the modules into memory and start
using the drive. There is no need to change the operating system when
the system hardware changes.

Most modules contain program code or constants. Data modules are an
exception. They can be created dynamically (All other types of modules
must be loaded from a disc.), and they are used to hold variable
information. They are excellent tools for inter-process communication
and medium term storage. Data modules stay in memory until they are
unlinked or the machine is turned off.

Input/Output

As far as possible, CD-RTOS has a consistent interface to all I/O devices.
A program needs to name a file or device to open it, but the open returns
a path number which is all the program needs for subsequent operations
on the file. A program can get the the details about an I/O path if it
requires them, but in many cases there is no need to know even whether
output is going to a screen, a printer, or a floppy disc drive. Sometimes
this is only a small convenience to programmers, but it can be
wonderfully useful.

Directories

CD-RTOS is able to arrange files on random access devices in
hierarchical directories to any depth. This is conceptually a simple idea,
though the standard manila folder analogy becomes brutally stretched by
the unlimited size and depth of directories. A disc, or any other random

203

204 CD-I: A Designer's Overview

access device, has a root directory which is always created when the disc
is formatted. The root directory and every other directory contains a list
of files and directories. You can name a file by giving the path list that
reaches it.

Typing long path lists for every file would quickly make directories an
unattractive organizational tool. The main CD-RTOS file systems
support two default directories, one for executable files, another for all
other files. Any path name that does not start with a slash (indicating a
device name), will be taken relative to a default directory. If the file is
opened for execution (as the load command would do), the execution
directory is used. If the file is not opened for execution, the default data
directory is used.

THE PARTS OF CD-RTOS

The base of the operating system is the kernel. This module supports the
services that a sophisticated operating system cannot do without. Input
and output are not handled by the kernel. It routes I/O operations to
operating system modules called file managers.

File managers typically deal with device-independant aspects of I/O. The
key file managers in CD-RTOS are CDFM and UCM. CDFM manages
files on compact discs, and UCM (User Communications Manager)
manages the CD-I user interface.

Other file managers from OS-9 are available for use in an extended
player. There’s a file manager called RBF (Random Block File Manager)
which manages files on mass storage devices like hard discs, floppy
discs, and RAM discs. Another file manager called SCF (Sequential
Character File Manager) handles devices like terminals, modems, and
printers. Nfm (Network File Manager) can be used to build a network of
CD- RTOS and OS-9 computers.

File managers can be independent of the details of the hardware they use
because there are other operating system modules called device drivers
that handle the details of the I/O hardware. RBF doesn’t care whether it
is handling a floppy disc, a hard disc, or a RAM disc. All those details
are handled by device drivers.

Device descriptors tie aname toa device driver, a file manager, and some
constants. Programs give the operating system the name, and the
operating system hooks the parts together to give the user the right file
manager, device driver, and the address for the I/O hardware.

Appendix C: Introduction to CD-RTOS and Invision 205

PROGRAMMING FOR CD-I

The CD-I File Manager

A Review of the Hardware

To understand the CD file manager (CDFM) you need to understand
some facts about discs and disc players. The low- level format of a CD-I
disc and the mechanism that plays it are are exactly the same as for
CD-audio. Because of this similarity CD-audio discs can be played on a
CD-I player, and CD-DA tracks from a CD-I disc can be played on a
CD-audio player.

The technology used for CD-audio was intended to offer about the same
functions as a record player. Random access to tracks of a recording, or
maybe a little finer, was sufficient. Delays of about a second when the
player skipped tracks were acceptable. Since CD-audio was designed
with these limits in mind, and CD-I was designed to be compatible with
CD-audio, it’s not surprising that it’s difficult to make disc players with
fast random access.

Some CD players seek faster than others, but the data rate of all CD
players is fixed by the specification. This is one of the ways that a CD-I
player is exactly the same as a CD-audio player. The player reads the
disc just fast enough to feed CD-DA information to an audio processor.
There is no way to increase the data rate on a CD-I player, and the only
way to decrease it is to ignore some of the data.

Let me put the performance of a compact disc in computer terms. For
simple reading, a compact disc player delivers data about five times as
fast as a disc drive, but any substantial seek is about ten times slower
than for a disc. A CD-I player can read the disc briskly but only for
sequential access.

Playing Real Time Files

CD-I type-A audio must be delivered to the audio processor half as fast
as it can be read from the disc. This gives some unused bandwidth for
other data, but only if the data can be reached without a seek.

The best way to arrange data for fast access by any disc-like system is to
group the data close together on the disc. A CD-I system makes it easy
for the disc designer to group related data. A single file can contain
computer data, sound, and video. The CD-I hardware and CD-RTOS can

206 CD-I: A Designer's Overview

separate the different data streams and deliver them where they are
needed.

Real-time data is information that must be processed quickly. CD-I
allows disc sectors to be tagged as real-time data. Though any data may
be marked with the real-time attribute, it is intended for time sensitive
data like audio codes. Because the real-time attribute is attached directly
to each real-time sector, files can contain ordinary data and real-time data
mixed in any convenient way.

The real-time sections of a file are called real-time records. CDFM will
treat real-time records as ordinary data if it is asked to read them, but if
a CDFM play command is used instead, the real-time attribute will be
honored.

The CDFM play commands are unlike any operations on a conventional
file system. A play command opens the flood gates and lets data from
the disc pour into memory as fast as the hardware picks it out. The
program that issued the play command can direct the stream of data, but
the only way it can effect the data rate is by changing the rules used to
select data from the disc. The program must keep up with the data.

The program has little control of the flow of data, but the rate can be set
when the disc is designed. If the designers only put a stream of monaural
level-C sound into some part of a real-time record, the data rate will be
only one sixteenth of full flow. The authoring system will arange the data
on the disc so the system can pick out each sector at the right moment,
and CDFM will see to it that they are delivered to the application or the
audio processor almost instantly.

The disc designers could have tucked other data between those audio
sectors. The play command divides the data flow from the disc into four
streams. Three streams go to blocks of memory provided by the
application. These streams are for audio, video and program data. The
fourth stream goes directly to the audio processor. The data on the disc
is tagged so CDFM will know where to direct it.

Each logical record in a real-time record has a channel number that was
assigned to it when the disc was designed. The CD-I system supports 32
channels of which 16 can be used for any type of data. The other 16
channels may be used for anything but audio data. CDFM can be directed
to select logical records from a file based on their channel numbers.

Appendix C: Introduction to CD-RTOS and Invision 207

CD-I channel numbers are not like television channels. A program need
not select a channel for a read or play command. It can choose to receive
any bundle of channels, even all of them.

Careful assignment and subsequent selection of channels can give a
single real-time record many different appearances. There are
4,294,967,295 different ways to choose channels - plus one if we count
choosing no channels at all.

The set of channels (the channel selection mask) can be changed between
real-time records. Since a real-time file can be made up of many real-time
records, the number of ways a real-time file can be played is actually
much greater than four billion.

Probably the simplest thing to do with a compact disc is to play strictly
sequential audio data. This can be done almost without program
involvement. The three streams of data that CDFM buffers for the
program can be empty or ignored. The only active stream is from the disc
to the audio processor.The program to play a disc with a couple of hours
of music could be as simple as an open for the music file and a play
command.

The three streams of real-time data that the play command directs to
memory are stored in buffers that are described by a play control block
(PCB) and three play control lists (PCLs). The play control block
contains control information for the play command and pointers to the
three play control lists.

The play control lists contain the detailed information that CDFM uses
to manager buffers. Each entry in a play control list contains control and
status fields, a pointer to a buffer, a pointer to the next play control list
entry, and a signal number that CDFM will send to the program if it
encounters a trigger interrupt on the disc or when it fills the buffer.

A program may fail to process a buffer in the interval between the time
CDFM fills a buffer and the time the next sector is ready. When a buffer
is full CDFM will move to the next play control list entry. If there isn’t
a buffer ready for the data, CDFM will send the data directly to the output
processor for that type of data (this really only makes sense for audio
data). It can’t wait for the program to empty a buffer. To avoid this
problem, a programmer should put two or more entries in each play
control list. With at least two entries in the list, CDFM can be filling one
buffer while the program is processing the contents of another buffer.

208 - CD-I: A Designer's Overview

The arrangement of links in a play control list is entirely up to the
programmer, but linking the entries into a loop should work well. The
CDFM will work its way around the links filling buffers while the
program follows behind.

When a play command is issued the file manager starts an asynchronous
activity. The activity is an operating system service on behalf of the
program - not a process. A program can control the play command
through the play control block, but the actual operation takes place inside
CDFM.

The play in progress can by monitored through a control block that
contains the current position in the file, but it’s more in keeping with
CD-RTOS style to give the play command some signal numbers so the
file manager will be able to notify the calling program when buffers are
filled, when a "trigger" in the disc file is encountered, and when the
command terminates.

The most extreme use of the asynchronous nature of the play command
is to start the CD-I file manager playing a file then ignore the progress
of the command. This would be the best way to play a file of strictly
audio information routed directly to the audio processor as was discussed
earlier. While data travels from the CD to the audio processor the
microprocessor is free to do anything that doesn’t directly involve the
audio processor or the disc. While the file plays, the microprocessor
might run a graphics display or even a text editor.

CD-DA Files

CD-DA is simple 16-bit digital audio. It is the highest fidelity audio
encoding method available to CD-I designers, but it uses the entire
bandwidth from the disc. Since no other data can be mixed with CD-DA
data, the disc is inaccessible while CD-DA is being played. This limits
the microprocessor and the video processors to whatever data is in RAM
before the play command starts.

A special CDFM command plays CD-DA files. The command always
routes data directly from the CD to the audio output. The only connection
that the CD-DA play command makes between a running CD-DA play
and the calling program is a signal that can be sent by CDFM when the
play is done.

There are a few brute force ways to influence a CD-DA play in progress.
The play can be paused or the disc can be ejected. The position in the file
can be determined to within a second by using a getstat system call. This

Appendix C: Introduction to CD-RTOS and Invision 209

is enough to implement the functions that most CD-audio players offer,
but no more.

Reading CD-I Files

In many ways CDFM makes CD-I files seem like RBF disc files. They
have a directory hierarchy, and files can be opened and read in random
order. Since CDFM ignores the contents of a file when it is processing a
read request, read is a good way to get simple data off the disc. Reading
is particularly appropriate for files that contain only computer data, but
files containing audio and video information can be treated like
conventional files when that is called for.

The User Communication Manager

The User Communication Manager, UCM, is the CD-RTOS file
manager that handles the four devices that are used primarily for
interaction with the user: video, audio, keyboard, and pointing device.
The connection to the keyboard and pointing device is fairly ordinary.
The connection to the audio output is moderately unusual, and the
interface to the video output can be complex and wonderful.

Keyboard Input

A read request to a UCM path will return characters typed at the
keyboard. A program can either issue a read and wait for input or ask
UCM to send it a signal when keyboard input arrives.

Since CD-RTOS will buffer keyboard input whenever it is entered, a read
request might be satisfied immediately with data in the keyboard input
buffer. A program can ask UCM how much data is waiting in the buffer
and use that information to decide whether to simply read from the
keyboard or ask to be signaled when more data arrives.

The Pointing Device

The simplest way to use a pointer device is to use the PT- Coord function
to get the current coordinates of the pointer. If an application needs to
track the pointing device, it must poll UCM for the current location
frequently enough to respond smoothly to any movement.

Polling steadily for the pointer location is wasteful. Getting the
coordinates twenty times per second might be enough to track the pointer
responsively, but it would be useless except when the pointer is in

210 CD-Il: A Designer's Overview

motion. CD-RTOS lets a program track the pointer without wasting time
polling when the pointer is at rest.

The program can ask UCM to send it a signal every time the pointer
location changes. A program that uses this facility need only ask for the
pointer coordinates only when they change. This prevents the program
from wasting time checking the pointer when it is not moving. It also
adapts the rate at which the program checks the coordinates to the speed
of the pointer. The program will check the pointer exactly often enough
to track it accurately.

Audio Output

Audio output revolves around soundmaps. A soundmap is a data
structure that UCM uses to store data that can be routed to an audio
processor. The information in a soundmap is encoded in sound groups,
just as it is stored on a disc. Internally, a soundmap is a block of sound
groups, the number of sound groups in the block, and a code that indicates
the way the sound is encoded.

A UCM function creates soundmap data structures. It returns a pointer
to a soundmap which can be used to load the structure with audio data.
It also returns a soundmap identifier which is the name by which UCM
recognizes the soundmap.

All UCM sound-manipulation operations are performed on soundmaps.
UCM will also output soundmaps (and only soundmaps) to the audio
processor of your choice.

Two UCM instructions merge a pair of soundmaps into a third soundmap.
Two monaural soundmaps can be combined into a stereo soundmap, or
two stereo soundmaps can be mapped into another stereo soundmap.
These operations don’t perform any actual mixing of the sounds
represented by the soundmaps (in the sense of an audio mixer). They only
copy channels intact from soundmap to soundmap.

Ordinarily the length of time a soundmap will take to pass through the
audio processor is determined by the audio coding method and the
number of sound groups in the soundmap. A programmer can lengthen
this duration vastly by installing loops in the soundmap.

A UCM function can put a loop in a soundmap by giving the number of
the first and last sound group in the loop and the number of times the
loop should be submitted to the audio processor. A loopback could be
used to extend a uniform sound for a specified interval, or to repeat a

Appendix C: Introduction to CD-RTOS and Invision

bounded sound a specified number of times. The sound of a basketball
being dribbled could be made by looping around the sound of one
bounce.

The features of the audio loopback are not yet precisely defined. It might
be possible to use several loops in one soundmap. Perhaps a roar of
applause followed by rythmic clapping and finally a dying patter of
individual claps could be compressed into an applause loop, an
intermediate section where the rythm begins, a rythm loop, and a dying
out section with many little loops. More complex rythmic effects might
be made with sequenced and nested loops.

There are two UCM functions that can be directed at audioprocessors
without mentioning soundmaps. One function amounts toa volume, pan,
and balance control. It controls the attenuation of each path in the audio
processor: left to left, right to right, left to right, and right to left. An audio
processor can also be turned off - thereby interrupting any soundmap that
is active.

The main source of soundmap data is the CDFM play function, though
programs are free to generate soundmap data. Generating or
manipulating soundmap information is difficult. It would be an
ambitious project to generate or modify ADPCM codes in software, and
soundgroups must be stored as ADPCM data.

There is a special case of ADPCM data that is comparatively easy to
work with. If the filter coefficients in type-A ADPCM soundgroups are
set to zero, the result can be manipulated like 8-bit PCM data.

Video Output

Drawmaps are the UCM data structure for video output. They can be
loaded by a CDFM play command or under program control. A wide
selection of operations on draw maps is supported by UCM. These
operations can combine drawmaps in many ways, draw characters or
graphics, and display the drawmaps with great flexibility.

Text

UCM supports multiple text fonts and supplies several tools for writing
text into a drawmap. These tools are divided into two groups: one treates
a drawmap like a video terminal, the other uses a drawmap as the bit
mapped object that it is.

211

212 CD-I: A Designer's Overview

Any program that works with a video terminal can be converted to use
UCM’s terminal emulation mode. The emulation supports all the features
of a simple terminal including line editing and character attributes. The
screen is controlled by writing text and control codes to it with the
CD-RTOS I$Write command. UCM terminal emulation is not the best
way to use the full power of CD-I video, but it is a familiar interface with
enough power for many applications.

UCM also provides tools that draw text on the screen. These commands
take a precise location and a text string. The text string is drawn on the
screen with the baseline of the first character at the specified location.
One of the text drawing commands will justify the string to a given
length. Control codes are not used with the text drawing commands. This
leaves more work for the program, but it lets characters be placed at the
exact locations the program specifies. The terminal emulator only allows
positioning to a line and column.

Fonts

Extensive support for text fonts is built into UCM. Fonts are described
by data modules that must be in memory when the fonts they describe
are in use. The system can have up to 4 fonts active at any time with a
total size of up to 64K characters.

A character is always designated with a single code, not a font number
and a character number, nevertheless each character code selects a font
as well as a character. The way a character code selects a font and a
character within the font depends on the active fonts and the way UCM
interprets the code.

UCM has three modes modes for interpreting character codes. Eight-bit
codes range from 0 to 255 - enough for most character-based languages
- maybe enough for a few type styles. Seven/fifteen-bit codes represent
128 characters in eight bits. The high-order bit is reserved as a mode
switch. If the high-order bit is on, 16 bits are used for the code. The
seven/fifteen-bit method generates character numbers zero through 127
with eight bits per character, and 32768 through 65535 with 16 bits per
character. This gives enough codes for oriental languages and many type
styles. Careful assignments of character codes will let frequently used
characters be represented with the 7-bit codes. Sixteen-bit codes don’t
offer the compression opportunity of seven/fifteen-bit coding, but
sixteen-bit coding is simple and it can represent 65536 characters.

If one font can contain as many as 65536 characters, why does UCM
bother to support four fonts at a time? The reason for multiple fonts is

Appendix C: Introduction to CD-RTOS and Invision 213

tied up with the way fonts are coded. Each character in a font has a
number, a bit map, and a width. All the characters in a font share height
characteristics, a proportional/monospace attribute, and a coding
method. To get variety in some attributes (like character height) an
application must use several fonts.

Since UCM does not use a separate font code for each character, each
font/character pair must have its own number. This requires some care
when numbers are assigned to characters in a font (which is done when
the font module is made). UCM will not tolerate overlapping codes
between fonts.

Operations on Drawmaps

The User Communication Manager has a variety of tools for copying a
section of one drawmap into another drawmap. The simple tools copy or
exchange rectangular regions between two drawmaps. Other tools offer
variations on the simple ones.

There are special copy and exchange operations that accept a transparent
color. These operations will not copy pixels that represent the transparent
color. The pixels that would have been replaced by the transparent color
are unchanged by these operations. This gives UCM the ability to
simulate some of the transparency operations supported by the video
hardware.

Rectangular areas or individual pixels can be copied in and out of a
drawmap. The UCM converts the image data into a standard array of
pixel information when it reads it out of a drawmap, and converts it from
pixel form to the drawmap’s internal form when it writes data into the
drawmap.

An irregular write operation permits updates of selected parts of a
drawmap. The operation updates each line in a given range, with the ©
extent of the update specified on a line-by-line basis. The update for each
line has a location and a length. Lines that don’t need updating can have
a zero-length update.

The operations on drawmaps can be used to conserve disc bandwidth.
When a portion of an image must be updated, that portion can be read
(or played) from the disc and used in an drawmap-update operation. In
some cases partial updates can give the appearance of full-screen
full-motion video.

214 CD-I: A Designer's Overview

The Graphics Cursor

The video processors support a graphics cursor, a 16x16 image plane that
sits in front of the other planes. The UCM offers a high-level interface
to the graphics cursor hardware. It includes functions to control the
existence, location, shape, color, and blinking of the cursor.

The conventional use of a graphics cursor is as an on-screen analog to
the pointing device. As usual, CD-RTOS does not enforce this
convention. Any use for a mobile 16-bit-square image plane is valid for
the graphics cursor. Perhaps it would be an easy way to represent icons
when they are in motion?

Clipping Regions

Part of adrawmap can be givena name. The part is selected and the name
assigned by creating a region. If two or more regions are defined in a
drawmap they can be combined with a region intersection, union,
difference, or exclusive-or operation. A region can also be moved within
a drawmap. Regions can used as clipping regions or they can be drawn
on the drawmap.

When a clipping region is "set" in a drawmap no drawing operation will
be allowed to change the drawmap outside the clipping region. This is a
useful feature for window support and graphics operations. Notice that
clipping really clips. Things that are clipped off are not stored anywhere.
If the clipping region is changed, the image will not be updated. (If you
need clipped data retained use a mask.)

Drawing Commands

The User Communication Manager offers a good set of graphics
functions. The scope of the graphics support is typical of a good graphics
package.

A Partial List of Drawing Commands

Set Drawing Pattern
Set Pattern Alignment
Set Color Register
Set Clipping Region
Set Pen Size

Set Pen Style
Draw a Dot

Draw a Line

Appendix C: Introduction to CD-RTOS and Invision

Draw a Polyline

Draw a Circular Arc

Draw an Elliptical Arc

Draw a Rectangle

Draw an Elliptical-Corner Rectangle
Draw a Polygon

Draw a Circle

Draw a Circular Wedge
Draw an Elipse

Draw an Eliptical Wedge
Draw a Region

Bounded Fill

Flood Fill

Copy Drawmap to Drawmap

Displaying Drawmaps

UCM does not have a function that simply displays the contents of a
drawmap on the screen. The program must provide a display control
program (a program that will be executed by the video processor). In its
simplest form the display control program directs the video processor to
display a range of lines from a drawmap. When its full power is used, a
display control program (DCP), can operate the video processor as a
special effects device.

One part of a display control program is executed every time the video
processor starts to display the screen (50 or 60 times per second); this is
called the FCT (Field Control Table). Another part of the display control
program is executed for each scan line; this is called the LCT (Line
Control Table). The same codes are used for the FCT and LCT. They
differ only in when they are executed.

Instructions placed in the FCT will be executed once each time the
display is refreshed. The FCT must include an instruction that points the
display processor to the LCT of the line that should be displayed at the
top of the screen. If there are any display other attributes that can be set
for the entire screen, the FCT might be a good place to set them. For
instance, the selection of the top image plane might be done here.

There must be a DCP for each of the two image planes (or paths). Each
path controls the parameters that effects it, but path 0 has a few extra
instruction codes that control both planes; e.g., plane order.

UCM supplies a series of commands that manage the components of a
DCP. These commands give programs a uniform interface to the DCP,

215

216 CD-Il: A Designer's Overview

but the application program is responsible for composing and modifying
the DCP programs.

The DCP instructions that link LCT to FCT, and LCT to LCT are the
only instructions specifically supported by UCM. These instructions are
useful for scrolling and split screen effects.

Most special effects are performed by manipulating DCP programs. A
cut from one image to another can be done by changing the top plane in
the FCT. A wipe can be done by changing the top plane one line at a time
or by using a matte. A fade can be done by decreasing the image
contribution factor of both planes over time. A dissolve involves
increasing the contribution of one plane while decreasing the other.

Display Parameters Controlled by the DCP

Image Coding Method

Transparency Control

Plane Order

Backdrop Color

Transparent Color (for each plane)

Mask Color (for each plane)

Image contribution factor (for each plane)
Mosaic Control (for each plane)

Color Lookup Table (for each plane)
Matting (for each plane)

DCP instructions can control video parameters on a line by line basis,
but they can’t directly affect anything other than the video processors.
For instance, there is no DCP instruction to read the disc or modify a
drawmap. There is, however, a DCP instruction that causes a signal to
be sent to the responsible program. These signals can be used to
synchronize program activity with DCP activity. UCM exposes the
power of the video processors to application programs by making display
control programs the responsibility of programs. This does not mean that
every program that displays video data must include explicit code that
manages the DCP. InVision provides a layer over most of UCM, hiding
many of the details of screen management.

The Real Time Record Interpreter

Real-time records can contain a RTCA (Real Time Control Area). These
control areas contain computer data that describe the operations that
should be used to process the data that follows. The language used in the

Appendix C: Introduction to CD-RTOS and Invision 217

RTCA is not as powerful as direct use of the UCM, but it is a convenient
way for each record to encode directions for its own processing.

A trap handler named RTRI (Real Time Record Interpreter) interprets
the codes stored in the RTCA. An application that wishes to use the RTRI
must open a real-time file and call the RTRI to play a specified number
of real-time records. The RTRI will handle the details of playing the file
asynchronously.

RTRI Functions

Soundmap Functions
Create Soundmap

Output Soundmap

Stop Audio Processor
Conceal Soundmap Error
Close Soundmap

Mix Mono to Stereo

Mix Stereo to Stereo

Set Soundmap Loopback
Set Attenuation

Drawmap Functions

Create Drawmap

Set Drawing Origin

Copy Drawmap to Drawmap
Exchange Between Drawmaps
Copy with Transparency
Exchange with Transparency
Irregular Write

Read Drawmap

Write Pixel

Read Pixel

Conceal Drawmap Error
Close Drawmap

Graphics Cursor Functions

Position Graphics Cursor
Show Graphics Cursor
Hide Graphics Cursor
Graphics Cursor Pattern
Graphics Cursor Color
Graphics Cursor Blink

218 CD-Il: A Designer's Overview

Clipping Region Functions

Create Region
Region Intersection
Region Union
Region Difference
Region Exclusive Or
Move Region

Delete Region

Drawing Functions

Set Drawing Pattern
Set Pattern Alignment
Set Color Register

Set Clipping Region
Set Pen Size

Set Pen Style

Set Transparent Color
Draw a Dot

Draw a Line

Draw a Polyline

Draw a Circular Arc
Draw an Elliptical Arc
Draw a Rectangle
Draw a Rounded Rectangle
Draw a Polygon

Draw a Circle

Draw a Circular Wedge
Draw an Ellipse

Draw an Elliptical Wedge
Draw a Region
Bounded Fill

Flood Fill

Draw from Drawmap

Fonts and Drawing Text

Draw Text

Character Code Mapping
Get Font

Activate Font

Deactivate Font

Release Font

Draw Justified Text

Appendix C: Introduction to CD-RTOS and Invision

Display Control Program Functions

Create FCT

Read FCT

Write FCT

Read FCT Instruction
Write FCT Instruction
Delete FCT

Create LCT

Read LCT

Write LCT Read LCT Column
Write LCT Column
Read LCT Instruction
Write LCT Instruction
Delete LCT

Link LCT to FCT
Link LCT to LCT
Execute DCP

Drawing Information Functions

Set Interlace Mode
Calculate Text Length
Relative Char. Positions
Return Font Data
Return Glyph Data
Justified Char. Positions
Is Pointer in Region
Get Region Location

Terminal Emulator Functions

Write

WriteLn

Set Output Drawmap
Set Mapping Mode
Activate Font
Deactivate Font

Pointing Device Functions
Get Pointer Coords.

Signal on Pointer Change
Release Device

219

220 CD-I: A Designer's Overview

Keyboard Functions

Read Line

Read

Signal on Data Ready
Release Device

Check For Data Ready

Functions for Reading the Disc

Load Soundmap

Load Drawmap

Read Data

Play to Audio Processor
Seek to Block

Housekeeping Functions

Read RTCA

Set Alarm

Set Channel Mask
Exit

Delete Label
Open Path

RTRI instructions include an opcode, control information, and
parameters. The opcode specifies one of the functions in the RTRI
function table or an application-specific function. These instructions are
executed by the RTRI in response to signals.

Instruction Map
S. Signal |T. Signal Sig Stop |Parm Length

Op Code Selects a RTRI function

S. Signal Signal to send on completion of function
T. Signal Signal to trigger this function
R. Count Kill this instruction after it has executed R count times.

Sig Stop Stop activating instructions for the trigger signal after this
instruction.

Appendix C: Introduction to CD-RTOS and Invision

Parm Length Length of the instruction’s parameter area
Parms The parameters for this RTRI instruction

Every RTRI parameter has a length and a label. The length is for the
parameter’s initializing value. If the length is zero the function has no
initializer. If the label is zero, the value of the parameter is not available
except to the function executed by this instruction. If the label is non-
zero, the parameter is shared among all the instructions refering to that
label number.

The flow of control in an RTRI program is not like an ordinary program.
There is no sequence of instructions that will be executed. All the
instructions listen for their trigger signal. When the signal appears they
queue up to execute.

Signals come from all the usual CD-RTOS sources and from the RTRI
itself. Each instruction can send a signal when it completes.

Picture each instruction as an active entity. They listen for their signal
and move to the execution queue whenever they hear it. When an
instruction is through executing it might send a signal to call on other
instructions to execute.

The instruction-termination signals can be used to make a sequence of
RTRI instructions execute. The first instruction can send a termination
signal that triggers the second instruction, the second instruction can
trigger the third, and so forth. Meanwhile any signals from outside RTRI
can trigger their own instruction sequences.

If an application needs a function that is not provided by the RTRI, the
function must be added to the RTRI instruction set for that application.
The application program should include code that implements the new
function. The function can be bound to an RTRI opcode by calling the
RTRI trap handler with the new opcode and the address of the
corresponding routine.

There is a speed penalty incurred by the RTRI (as with any interpreter).
When performance is the highest priority, ordinary programs can do
better than the RTRI.

The strong point of the RTRI is small programs. Its policy of loading the
instructions to handle each real-time record with that real-time record is
especially efficient.

221

222 CD-I: A Designer's Overview

Since RTRI instructions hide asynchrony to some extent, it may prove
easier to write correct programs for the RTRI than to write them in the
C language.

Non-Volatile RAM File Manager

A small amount of non-volatile RAM is included in the CD-I base case
hardware. CD-RTOS includes a special file system designed to have
modest storage overhead for the non-volitile RAM.

The non-volatile RAM file manager (NRF) is a simplified version of
RBF. The most noticable missing feature is directory support. There is a
single NRF directory which maintains a list of NRF files and their
attributes. Other NRF directories cannot be created.

CD-RTOS imposes no restrictions on the use of NRF files, but
programmers should respect the limited non-volatile storage on a base
case system.

USER INTERFACE FOR DEVELOPMENT SYSTEM

Applications on a consumer CD-I system may offer the user an interface
designed specifically for the application or they may use the InVision
user interface. InVision is a set of enhancements to UCM and a user
interface library that implements a loose standard for user interactions.

Designers are free to use InVision or modify it to give their applications
a special flavor. Provided that the flavor is not too special, designers and
users will be able to enjoy some variety and the benefits of a standard
interface.

THE PROGRAMMER’S VIEW

I have been discussing programming: in abstract terms. Unfortunately
there are some important issues that CD-RTOS programmers must
understand. This section will touch lightly on some important details.

Processes

CD-RTOS starts processes with a fork command or a chain command.
These differ in that chain replaced the current process with the new
process while fork creates a new process and leaves the parent process
running.

The parent process can specify the number of I/O paths that the child will
inherit, the contents of the child’s parameter area, an amount of extra

Appendix C: Introduction to CD-RTOS and Invision

memory for the child’s stack area, and the priority the child should run
at. The fork command returns the process number of the child.

When a process starts, almost all of the MPU registers are set to
significant values. Most important are the stack pointer, the PC, and a
Static storage pointer. Many operating systems don’t give new processes
a static storage pointer. Since CD-RTOS is a multitasking operating
system which does not require memory management, programs must be
able to run anywhere in memory. The operating system tells them where
it placed there static storage, and the programs take it from there.

Processes can have considerable influence on the way they are
scheduled. They can instruct the operating system to ignore them for a
length of time with a sleep command, they can wait for a child process
to terminate with a wait command, and they can change their scheduling
priority with a SPrior command.

Memory

One symptom of multitasking (without memory management) is that a
region of memory can be cornered by memory allocated to other
processes. When that happens to a program’s stack area, the amount of
the stack memory can’t be increased until the other process releases the
memory that is blocking the stack. CD-RTOS has a system call to change
the size of the stack allocation (like a Unix sbrk), but this call will fail if
memory contiguous with the stack allocation is not available. Another
system call that allocates blocks of memory wherever it can find them
should be used whenever possible.

Synchronization

The CD-RTOS file managers make extensive use of signals to
communicate with programs. Signals are the software equivalent of
interrupts. When a signal is sent to a process, the normal execution of the
process is interrupted and control is transferred to the process’s signal
handler.

A process can give CD-RTOS the address of a function which the
operating system will call each time a signal is directed at the process.
The signal function will be entered with the signal number as a parameter,
and the A6 address register set to a value that was specified when the
signal function was identified. By convention this register is the base
address for static storage, but an assembly language programmer could
use it for something else.

223

224 CD-I: A Designer's Overview

A process can mask signals that it doesn’t wish to receive. This is done
automatically when a process is in the signal function. Signals that are
not received are queued for later attention. A process that doesn’t wish
to handle signals can elect to mask signals, or to catch and ignore them.
If it does neither, any non-zero signal will kill the process.

Subject to some security restrictions, processes can send signals to one
another. Inter-process signals work exactly like signals from file
managers.

Where Signals Come From

UCM keyboard interrupt
UCM keyboard kill

UCM keyboard data ready
UCM mouse motion
UCM FCT signal

UCM LCT entry signal
UCM Done with soundmap
CDFM End of play

CDFM Play buffer full
CDFM Play hit a trigger

Other programs application-specific
CD-RTOS kernel Alarm signals

CD-RTOS supports another synchronization method called events.
Events have names and values and are accessible to any process that
knows their name. A process can wait for an event to reach a certain
value, or a value within a range. CD-RTOS commands set or increment
the value of an event. It is simple to use events as semaphores, and with
some imagination they can address almost any synchronization problem.

The primary job of a program in a CD-I system will usually be routing
real-time data. Since this is a signal-driven activity, programs are likely
to be bundles of functions and processes written for speed. Writing this
type of program is not like ordinary programming. Debugging
signal-driven programs is just nasty.

Modules

Programs for CD-RTOS must be position independent. The operating
system supports multitasking, and gives modules no way to specify
where in memory they should be placed. A program must assume that it
could be located anywhere in memory and that its variables could be
located anywhere else. Position independant code can be written for

Appendix C: Introduction to CD-RTOS and Invision 225

68000-compatible processors, and compilers for CD-RTOS will
generate correct code so far as they are able, but programmers should
keep inthe back of their minds that code and data can be at different
locations each time a program runs.

Whenever possible CD-RTOS programs should be re-entrant. Once a
module is in memory it is convenient for the operating system to let every
program that needs the module use the same copy. This is possible
provided that the code and constants in the module are not modified by
any process that uses it.

The Role of Programming

Most CD-I systems will probably look more like stereo systems than
computers. But they are computers - powerful ‘computers. A CD-I
application can include a tremendous amount of data, but the data will
be presented to the user by a program. It is probably best if users forget
that their CD-I appliance is a computer, but a CD-I designer must not.

No household appliance (except, perhaps a telephone) has had anything
close to the computing power of a CD-I system. We can only guess how
the tool will be used.

INVISION

INTRODUCTION

InVision is an example of an object-oriented multi-media user interface
designed for Compact Disc-Interactive (CD-I) players. Other interfaces
will be available in due course. Since InVision is intended for a consumer
product, it is designed to be simple and easy to use by the user of a CD-I
player. For the content provider, not only does InVision provide access
to standard UCM functions, it adds several powerful new capabilities to
the CD-I environment.

ARCHITECTURE

The diagram shows the base case CD-RTOS software modules and the
InVision modules and their relationships to each other. The thick lined
boxes represent InVision modules while thin lined boxes represent
CD-RTOS modules.

InVision consists of three modules, a Display Manager, a Presentation
Support Library, and a Visual Shell. The Display Manager isa CD-RTOS

226 CD-I: A Designer's Overview

subroutine module for the UCM file manager and provides access to the
video, audio, pointer and keyboards drivers like the User Communi-
cations Manager. It also manages screens to allow more than one appli-
cation to run at a time and action regions for simplifying input.

CD-I
Application
Presentation Support Real-Time Record
Library Interpreter

Display User Communications Compact Disc
Manager Manager File Manager
Keyboard Video Pointer
Driver Driver Driver

The Presentation Support Library is a CD-RTOS trap library. This
collection of subroutines further simplifies the task of manipulating the
display and obtaining input from the user. The Presentation Support
Library provides requesters, which are standardized ways of asking
questions of the user and controls, which are standard ways of getting
switch and volume control input from the user. The Presentation Support
Library includes a set of visual effects functions, a set of international-
ization functions to assist applications in being country independent, and
functions to maintain and determine certain preferences of the user.

The top level of InVision is the Visual Shell. The Visual Shell is the only
part of InVision that the user will actually see. It essentially forms the
control panel of the CD-I player. The Visual Shell is fully customizable
by the player manufacturer. In addition to its main function of starting
the play of a disc, the Visual Shell can execute a set of small utilities
called accessories. Accessories perform functions such as setting
preferences, setting a clock, or controlling other parts of an audio-video
system. The accessories that are included with a CD-I player may also
be determined by the manufacturer.

Appendix C: Introduction to CD-RTOS and Invision

THE DISPLAY MANAGER

The InVision Display Manager (DSM) isaCD-RTOS subroutine module
to the UCM file manager and provides InVision’s interface between the
CD-RTOS kernel and the CD-I video, audio, keyboard and pointer
drivers. It is also the application’s basic interface to those drivers. In
addition to providing access to standard UCM functions, the Display
Manager introduces two new capabilities to the CD-I environment,
screens and action regions.

SCREEN MANAGEMENT

The Display Manager provides screens so that more than one application
can be in progress and providing information to the user of the CD-I
player. Since there is only one physical display for the system,
multi-tasking causes several applications to both attempt updates on the
same display. The Display Manager gives each of the applications a
*logical screen’ on which to display their output. At any given time only
the top screen is visible on the display and the application running on
that screen is the only interactive application available.

The screen management functions are based on the display control
program functions supported by the video driver. In addition to providing
access to the video hardware’s field control tables (FCT’s) and line
control tables (LCT’s) in much the same way that UCM does, there are
functions which coordinate the use of those elements in relation to the
logical screen.

Each screen appears as a single display control program to the application
that is using it. It is a collection of at least two LCT’s, one for each plane.
All of the functions for creating, deleting , reading, writing and linking
LCT’s provided by UCMare also provided by the Display Manager. This
part of the design of the Display Manager has been kept the same as UCM
so that applications have virtually the same ability to generate visual
effects, including the ability to pre-master the data for those effects.

Functions are also provided to associate a FCT’s with each screen. These
provide the ability to load instructions which affect the entire screen or
load a large number of color look-up table registers in the hardware.
When a screen is to be the one displayed its FCT’s are executed as the
current hardware display control program.

There are functions for showing, hiding, raising and lowering screens.
These functions are used to switch between screens when running
multiple applications. For example, a user could be interacting with one

227

228 CD-I: A Designer's Overview

application and want to access the Visual Shell for a moment. The Shell
screen will become the visible screen allowing interaction with the
Visual Shell. When the user us done they can return to the original screen
and interact with the application again.

— FI

Display Screen Stack

A diagram of a collection of screens is shown along with a diagram of
how the screens would appear on the display. The screens are maintained
in a stack mush like a stack of paper. In the figure, the individual screens
are the same size and are aligned with each other and the display. Since
screen C is at the front of the stack, it is the only one that is visible on
the display. :

The next diagram shows how the screen stack and display change when
anew screen is created and displayed. Whenever a screen is created it is
added to the top (front) of the stack.

Display Screen Stack

The last diagram shows how the screen stack and display change when
a screen (screen B) is raised. Screen B moves to the front of the stack
and becomes visible.

Appendix C: Introduction to CD-RTOS and Invision 229

Display Screen Stack

MESSAGE MANAGEMENT

The Display Manager provides a simple method of inter-process
communication. This facility is mainly used by the Display Manager to
transmit packets of information about pointer activity to applications. It
is general enough, however, that it may be used by applications using
InVision to send messages to each other.

The Display Manager provides each process using InVision with a queue
for these messages. Several functions are provided for manipulating the
message queue. The send function adds a new message to a queue. The
receive function retrieves a message from the queue. Specific types of
messages may be selected with the receive function. There is also a
function to determine if a message or a particular type of message is in
the queue. Particular types of messages may also be flushed from the
queue.

ACTION REGIONS

Action regions provide a mechanism for condensing and organizing the
continuous stream of coordinate and trigger data that comes from the
pointing device. By using action regions, applications can define areas
of the screen where the application is notified of the specific area where
the pointer activity occurred and the type of pointer activity that
occurred.

Action regions are based on the region mechanism provided by the video
driver. Whenever a screen is created a conceptual third plane, called the
action region plane, is created and placed in front of the two image planes
provided by the CD-I video hardware. This action region plane can also

230 CD-l: A Designer's Overview

be thought of as the cursor plane, since cursor movement is generally
associated with pointer movement.

The Display Manager continuously monitors the stream of data coming
from the pointing device, determines which screen and action region the
pointer activity is associated with, and sends a message to the application
associated with the screen and action region. The message contains the
action region identifier, coordinate, and type of pointer activity. Pointer
entering action region, pointer exiting action region, pointer moved,
trigger up and trigger down are the possible types of pointer activity.

The Display Manager provides a very powerful set of functions for
manipulating action regions.- The most basic of these is the move
function. Any action region may be moved to any position in the action
region plane. This allows the possibility that action regions might overlap
each other. To resolve these conflicts, the Display Manager maintains
the action regions in a stack. An action region higher in the stack has
priority over one lower in the stack. The create function adds new action
regions to the top of the stack and the delete function removes action
regions from the stack. The raise function moves an action region to the
top of the stack while the lower function moves one to the bottom. The
activate function causes the Display Manager to consider the action
region when it searches for an action region to be associated with a
pointer activity. The de-activate functions causes the Display Manager
to ignore the action region in its search.

Action regions are also hierarchically organized. Child action regions
may be created and associated with their parent. Whenever child action
regions are moved, they are positioned with respect to their parent. These
child action regions are maintained in stacks which are attached to the
parent, yielding a tree of action regions. When the Display Manager
searches the tree for an action region associated with a pointer activity,
it performs a top-down recursive search, looking for the highest priority
action window lowest in the tree which contains the coordinate of the
pointer activity. The action region functions manipulate the child stacks
exactly like parent stacks.

The hierarchical organization of action regions allows the creation of
groups of action regions which can be treated as one unit. Whenever a
function is performed on an action region, the same action is performed
on its children. If an action region is moved, all of its children move with
it. Raising, lowering, activating or de-activating an action region causes
the same thing to happen to all of its children.

Appendix C: Introduction to CD-RTOS and Invision

THE PRESENTATION SUPPORT LIBRARY

The Presentation Support Library is a collection of high-level functions
for simplifying application development for CD-I systems. Many of the
concepts found in traditional computer based user interface libraries are
included in the Presentation Support Library but are tailored to make use
of the multi-media capabilities of the CD-I system and oriented towards
a consumer based product. The main functional capabilities of the
Presentation Support Library are:

* Controls. Controls allow an application to obtain input from the user
using objects similar to switches and volume controls.

Requesters. Requesters allow an application to ask the user a question
and obtain a reply.

Visual Effects. Functions are provided to simplify the task of genera-
ting visual effects such as mattes, wupes, fades, dissolves and cuts.

Internationalization. Functions are provided for an application to se-
parate country dependent data such as prompts from the application
so that it may be country independent.

Preferences. Functions are provided for maintaining user preferences
such as language or nationality.

CONTROLS

Controls are a high-level mechanism for obtaining input data from a user
of a CD-I player. They provide applications with a mechanism to draw
objects on drawmaps and associate action regions with them. These
objects may also change appearance depending on the type of pointer
activity in the action region. When the sequence of pointer activity in the
action region required by the control occurs, an application defined
function can be executed to take action based on the value of the control.
The concept of controls in the Presentation Support Library is very
general to allow the application great flexibility in generating this type
of user interaction. An application may create its own control behavior
or use the standard control behaviors.

The most common types of controls are buttons and variable controls.
The Presentation Support Library provides three types of standard
control behavior, pushbuttons, switches and variable controls. The
application is able to customize these controls by providing the image
data for each of the states of the control. This makes it possible to have
controls built from natural images. The figure shows an example of a
screen for a CD-Audio player accessory. This accessory uses

231

232 CD-I: A Designer's Overview

pushbuttons for the play, pause, stop, fast-forward, rewind and eject keys.
Variable controls are used for the volume, balance, bass and treble
controls.

; CD Audio Player - ane }

Current Selection: Ploy Pause Stop

DH He

0 Volume jo FF Rew Eject

QO Balance 19 0 Bass 10 O Treble 19

Pushbutton type controls behave in a manner similar to a momentary
contact switch in that they are used to initiate a specific action by the
application. Pushbutton controls appear to the user in one of four states
depending on pointer activity in the control’s action region and the
application. The disabled state is set by the application and is used to tell
the user that the button will not do anything. When the button is not
disabled and the pointer is not in the control’s action region it is in the
normal state. When the pointer enters the control’s action region the
control enters the highlighted state, signifying to the user that the pointer
is on the button. If the user then depresses a trigger button, the control
goes to its active state. When the user releases the trigger button, the
control is returned to normal state and an application specified function
is executed.

Switch controls behave similar to an on-off switch. Switches controls
have a value associated with them, specifically, on or off. Switches
controls appear to the user in one of five states. disabled, normal on,
normal off, highlighted on and highlighted off. As with pushbuttons,
switches can be disabled by the application. When the switch is not
disabled and the pointer is not within the control’s action region, the
control is either the normal off or normal on state. When the pointer
enters the control’s action region it enters one of the highlighted states.
When the user depresses and releases a trigger button, the value of the
control is changed from off to on or from on to off. Whenever the value
of a switch is changed an application specified function is executed. The
new value is passed as a parameter.

The variable control is used to implement dials and slides and may have
one of many values associated with it. Visually, the image of a variable

Appendix C: Introduction to CD-RTOS and Invision 233

control has two parts, a background and a moveable part. The moveable
part appears to the user in one of three states, disabled, normal and
highlighted. As with pushbuttons and switches, the application may
disable a variable control to signify to the user that the control will not
do anything. When the control is not disabled and the pointer is not in
the control’s region, the control is in the highlighted state. If the user
depresses a trigger button, the user can move the moveable part of the
control by dragging. The application can specify two functions to be
executed by this type of control. One is executed while the button remains
down and the pointer is moved. The other is executed when the trigger
button is released. Both are passed the value of the control as a parameter.

REQUESTS

Requests, like controls, are a high-level mechanism for obtaining input
from the user of a CD-I player. Applications use requests to present a
question to the user of the CD-I player and get a response to a range of
choices. Requests are typically used at decision points in an application
where the user can guide its flow. They are similar to the pop-up of
pull-down menus supplied with computer based user interface toolkits.
While the application has complete control of the appearance of requests,
including the ability to supply actual image data, the question and choice
behavior of it is not changeable. The figure shows the CD-Audio player
accessory with a request on it, asking the user to select a song to play.

iui CD Audio Player ane ig

Pause Stop

C=)
Rew Eject

Ca)

Treble 149

[| until we have faces
re) *

Requests consist of a question and several possible responses. When the
request is drawn on a drawmap, an action region is associated with each
choice. Each choice can appear to the user in one of three states, disabled,
normal and highlighted. When the application disables a choice it signals
the user that the choice is not a valid one. If a choice is not disabled and
the pointer is not in its action region, the choice is in the highlighted state.
If a trigger button is depressed and released then an application specified
function is executed. It is passed the choice as a parameter.

234 CD-I: A Designer's Overview

VISUAL EFFECTS

The Presentation Support Library includes a number of functions to
simplify the generation of such visual effects as wipe, fade, dissolve,
matte and cut. The Field Control Tables and Line Control Tables which
control the display contain parameter information in the form of
instructions. Some of the visual effects functions generate lists of
instructions which are suitable for writing into LCT’s. One of these
functions takes a drawmap and some coordinate information and returns
a list of "load video start address" instructions which, when written to
and LCT, will cause the selected portion of the drawmap to be displayed.
Another takes a region created by the video driver and returns a list of
"load matte register" instructions for writing into anLCT.

The other visual effects functions perform timed writes to LCT’s. For
doing fades and dissolves, there is a function which automatically
generates a "set image contribution factor" instructions and writes them
into the LCT ata rate specified by the application. Other functions write
lists of instructions specified by the application on a timed basis.

INTERNATIONALIZATION

Since the CD-I market is international in scope, a significant number of
applications will need to be country independent. The largest area of need
for this is when applications prompt users for input using requesters or
controls and in the display of textual information. To help solve this
problem the Presentation Support Library provides resource modules.
Content providers can store country dependent information in these
modules. All an application need do is obtain the resource module for
the correct language and use the data in it. A small number of functions
are provided for properly formatting date, time and money strings
according to individual countries’ standards.

PREFERENCES

The Presentation Support Library maintains a file on the CD-I player’s
non-volatile RAM. It is mused to store the user’s preferences.
Preferences are simply parameters which affect the operation of the CD-I
system which can be adjusted by the user. One of these parameters is the
nationality of the user. This should be consulted by the country
independent applications so they can determine how to present
information to the user in a way that he is culturally accustomed to. Other
parameters are key delay time and key repeat time.

Appendix C: Introduction to CD-RTOS and Invision

THE VISUAL SHELL

The most visible part of InVision is the Visual Shell. The Visual Shell is
the first program that the user interacts with when he turns on the CD-I
player. The Visual Shell is designed to be the control panel of a CD-I
player. Its main purpose is to start the play of discs and start the various
accessories.

The Visual Shell is created using a request. The choices given in the
request correspond to the accessories that are available on the player. The
Visual Shell is fully customizable. When it is executed, the Visual Shell
accesses a resource module that the manufacturer has included in the
ROM. This module contains the image data for the background screen,
the data required by the request and the names of the individual
accessories that may be executed. The Visual Shell then waits until the
user either inserts a disc or selects one of the choices on the display using
the pointing device.

Small programs, called accessories are also provided with the Visual
Shell. The programs are used to set devices such as a clock or maintain
data in the preference file, or use the CD-I player as a CD-Digital Audio
player. What accessories are provided with a player is determined by the
player manufacturer. If a remote control bus is available on a CD-I player
it could become the centerpiece of an audio-video system. Accessories
could be provided that control each component on the bus.

Further information on InVision can be obtained from Microware
Systems Corporation, Des Moines, Iowa, USA.

235

INDEX

action area/region 49, 93
ADPCM 120

analog 16

animated figure 92
animation 32

application 49

application software 130
APU 72

audio 18, 23, 65, 70, 94, 96, 132,
139

audio control 26

audio pathway 132

audio processor 120
audio quality 23

audio sector 113
authoring environment 61

base case decoder 117
border color 128
brightness 38

budget 56

calendar 130

CD 15

CD-control unit 120
CD-DA 12, 15, 23
CD-drive 118

CD-I 3, 14, 15, 17, 21, 107
CD-ROM 3, 12, 15
CD-RTOS 18, 78, 133
CD-V 12, 15

CDFM 139

chroma-key 38, 102

clock 130

CLUT 19, 30, 33, 69, 75, 122
CLUT image 115

CLV 80

coding combination 122
compression 28

consumer 7, 21, 47, 49

Index 237

constraint analyzer 64
control 140

control device 22
CSD 135

cut 34

data acquisition 64
data format 64

data management 86
data sector 117
database 65
dataflow 77

DCP 126

decoder 117

design 66

design brief 45
design process 19, 46, 53
designer’s station 63
design team 54
digital 16

directory 111

disc building 64
disc label

disc space 67, 95
disc structure 109
display manager 142
dissolve 40

DMA controller 129
drawmap 42, 138
driver 140
dual-plane 81

DVI 4

DYUV 19, 29, 81, 123
DYUV image 115

electronic publishing 11
encyclopedia 85

fade 38

238 CD-I: A Designer's Overview

file 112

file manager 136

file protection 136
French phrasebook 100

golf 9, 22, 89
granulation 34

graphic control panel 87
graphics 29

hardware cursor 128
hyper-media 5

icon 95

idea map 52, 57, 59
image coding 81
image plane 31
image store 133
input device 51
intensity 38
interactive 6, 14, 43, 50, 74, 85,
99

interactive design 81
interface devices 74
interleaving 77
InVision 51, 142

kernel 134

L
links 88
M

magnification 36
matte 39, 127

menu 86

message sector 110
microprocessor 73, 92
mosaic effect 36

motion 32
multi-media 3
multi-media CD-ROM 4

natural images 29
NRF 140
NV-RAM 130

operating system 133
optical disc 12, 14
optical media 14
optical recording 13

path table 111
photographic database 89
physical interface 42
pixel hold 36

pixel repeat 36

plane effect 38

play 141

pointing device 130
presentation editor 65
presentation support library 142
production facility 63
production process 57
program module 67
prototype 57

Q
QHY 29
R

RAM 72, 130

RAM, non-volatile 130
real-time 141

real-time data 41

real-time dramatization 102
real-time interactivity 41
region 138

resolution 26, 68

RGB 19, 28, 29, 39, 116, 122,
125

RTCA 141

RTRI 141

run-length 124
run-length code 30,
run-length image 116

schedule 56

screen design 74
screen effect 75, 87
screen interface 86
screen update 97
script 64

scrolling 35

sector 67, 112

seek 97

seek time 79

shape recognition 99
simulator 66

single plane 34
soundmap 25
specification 109

standard 16, 26, 31, 64
start-up procedure 134

status descriptor 134
storyboard 58, 61
sub-screen 34
subsystem 62
surrogate travel 94

synchronization 41, 80, 140

testing 66
text 29, 31
tool 63
track 67

track organization 110

transfer path 130
transparency 38
trigger button 42

UCM 136
user interface 42, 139

video 18, 26, 32, 65, 69, 138
video pathway 132

video processor 121

video quality 27

video sector 114

visual effects 33

visual plane 19

visual shell 142

wipe 40
WORM 12

Index 239

The definitive explanation of Compact Disc — Interactive, written by the
authors of the system, Philips International of the Netherlands.

The book was written by a team of Philips and associated company
technical experts, and design professionals working in the field, all too
numerous to name individually. This book is essential reading for those
wishing to familiarize themselves with this new technology, and places

special emphasis on the design aspects for developing CD-I discs.

Compact Disc — Interactive, CD-I for short, is the latest development in
the field of Optical Recording, based on the highly successful
CD-Audio, those shiny silver discs that make that perfect sound.

CD-I starts from the existing capability of CD-Audio, and adds pictures,
cartoon animation, text, computer programs, as well as a whole range of
audio quality levels too. And as the name says, CD-I is interactive, in
other words, it responds to the demands of the user, and provides the
information he or she requests directly.

CD-I has been defined by the Compact Disc license holders,
N.V. Philips and Sony Corporation, as a complete specification,
covering not only how information is to be stored on the disc, but also
how the stored information is encoded and decoded.

This book provides full details on what CD-I can do, and how it works in
practice. A series of example applications are given to show the
designer how to approach the design of CD-I discs.

ISBN 90 201 21219

GIS

Taltsiesteltilis

CD-I A DESIGNER’S OVERVIEW Edited Niet mass. '

Pr |

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Kluwer - Compact Disc Interactive Designer's Overview - Second Edition"

See other formats